[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M71F43CZ9VW7C0K364JBAQJ1":3,"skill-same:01M71F43CZ9VW7C0K364JBAQJ1":236},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":16,"sourceHtmlUrl":17,"categories":18,"keywords":32,"publishedAt":52,"linkOnly":53,"markdownContent":54,"markdownBody":55,"frontMatter":56,"refName":57,"commitSha":58,"skillDirectory":5,"forksCount":14,"repositoryHtmlUrl":59,"repositoryHomepageUrl":6,"ownerHtmlUrl":60,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":61,"install":62,"resources":75},"01M71F43CZ9VW7C0K364JBAQJ1","editaplot",null,"单独科研数据绘图优先生成可手工微调的 Origin OPJU，无需点名 Origin；仿真实验自动出图和指定图源保持原方式。 Analyze local scientific CSV, TXT, XLS, or XLSX data; recommend publication-informed charts and Chinese scientific palettes; freeze a reproducible plan; automate editable figures through a callable local Origin\u002FOriginPro installation on physical Windows 10\u002F11 x64; and prepare or review user-edited Origin OPJU files. Use for beginner “drop in a file and draw it” requests; “我用 Origin 改过图了”, “审阅 OPJU”, “重新导出 OPJU”, “prepare-opju-review”, or “review-opju” requests; XPS, XRD, XAS, PL\u002FTRPL, DSC, NMR, FTIR\u002FIR, UV-Vis, electrochemistry, medical\u002FAI evidence, distribution, relationship, error-bar, bar, stacked, pie, Sankey, radar, heatmap, or verified 3D workflows; project-local Python setup; palette selection; and OPJU\u002FPNG\u002FPDF\u002FTIF verification. Do not use on macOS, Linux, WSL, Wine\u002FCrossOver, Parallels, or other VMs; to install or modify Origin; to redistribute reference images; or to claim an unverified Origin route.","SanAntonio021\u002Fagents-skills","SanAntonio021","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F56460680?v=4","01M71F2WGFWC4J9XYAKBEYGHSY","Weipeng Wang","01M71F2WHCP3JATT5FGJSXZXXC",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002FSKILL.md",[19,24,27],{"categoryNo":20,"categoryName":21,"taxonomyType":22,"taxonomyTypeDesc":23},"01JBSEEDCAT000000000000002","文档处理","SCENARIO","使用场景",{"categoryNo":25,"categoryName":26,"taxonomyType":22,"taxonomyTypeDesc":23},"01JBSEEDCAT000000000000004","电子表格",{"categoryNo":28,"categoryName":29,"taxonomyType":30,"taxonomyTypeDesc":31},"01JBSEEDCAT000000000000033","数据分析师","PROFESSION","适用职业",[5,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"local","relationship","drop","workflows","uv","files","electrochemistry","medical","claim","selection","project","xlsx","Charts","chinese","3D","PDF","distribution","error","images","2026-11-10 10:12:30",false,"---\nname: editaplot\ndescription: 单独科研数据绘图优先生成可手工微调的 Origin OPJU，无需点名 Origin；仿真实验自动出图和指定图源保持原方式。 Analyze local scientific CSV, TXT, XLS, or XLSX data; recommend publication-informed charts and Chinese scientific palettes; freeze a reproducible plan; automate editable figures through a callable local Origin\u002FOriginPro installation on physical Windows 10\u002F11 x64; and prepare or review user-edited Origin OPJU files. Use for beginner “drop in a file and draw it” requests; “我用 Origin 改过图了”, “审阅 OPJU”, “重新导出 OPJU”, “prepare-opju-review”, or “review-opju” requests; XPS, XRD, XAS, PL\u002FTRPL, DSC, NMR, FTIR\u002FIR, UV-Vis, electrochemistry, medical\u002FAI evidence, distribution, relationship, error-bar, bar, stacked, pie, Sankey, radar, heatmap, or verified 3D workflows; project-local Python setup; palette selection; and OPJU\u002FPNG\u002FPDF\u002FTIF verification. Do not use on macOS, Linux, WSL, Wine\u002FCrossOver, Parallels, or other VMs; to install or modify Origin; to redistribute reference images; or to claim an unverified Origin route.\n---\n\n# EditaPlot\n\n## 与论文图件规范的分工\n\n单独提出科研数据绘图请求时，在当前 Origin 路线能满足要求的前提下优先使用本技能，交付用户可继续手工微调的原生 OPJU；无需用户明确说 Origin。论文结果图由 `paper-figure-review` 确定技术含义、坐标单位和投稿排版要求，本技能负责导入、制作、保存和导出，再按该要求检查实际输出。\n\n论文任务已确定的字体、字号、尺寸等要求优先于下文及参考资料的 Origin 模板默认值；当前路线不能实现或回读验证时，报告缺口。跨技能继承已确定的任务输出目录，用现有输出参数传入，无需再次询问目录。\n\n科研示意图走 PPT、SVG 等设计路线。仿真、实验程序运行时的自动出图保留现有实现；用户指定 Python\u002FMATLAB 或沿用既有图源时遵循指定，不自动迁移。仅审查时给意见，不生成修改版。\n\n两个技能复用同一任务的数据含义、样式和授权，不各问一轮确认。用户保存后的当前 OPJU 是后续读取依据，按现有副本审查流程导出，不覆盖其手工调整；这不代表已提供任意 OPJU 原位编辑接口。\n\n制作前核对必要坐标尺度、阈值线、标记、版面与导出格式是否被当前路线支持；仅模板名匹配或 Origin 可启动不算满足。能力不足时说明具体缺口和可行替代，不静默把对数轴换成线性轴，也不把不符合要求的试图标为完成。\n\nTurn a scientific question and a read-only table into an auditable, editable Origin figure. Keep\nthe beginner experience conversational; use the deterministic engine for inspection, planning,\nrendering, exporting, and readback.\n\n## Request only scoped Windows permissions\n\n- Read the configured versioned runtime, selected table, and optional local reference image. Read the\n  complete repository only when working from an explicitly selected clone or approved source update.\n- Write only to the configured runtime's `.editaplot-venv`, an approved source repository during a\n  separately authorized update, and the established task output folder (defaulting to the selected\n  source file's parent folder). Never patch `.cc-switch`, `.codex`, or `.claude` Skill copies directly.\n- Run the local launcher, PowerShell\u002FPython subprocesses, and an EditaPlot-owned Origin instance in\n  the same active interactive Windows user session.\n- Use network access only for repository download\u002Fupdate and locked dependency retrieval. Treat a\n  user-scope winget Python installation as a separate system change that still requires explicit\n  consent.\n- Do not request administrator rights, mouse control, whole-drive write access, cloud upload of\n  private inputs, or DCOM, registry, firewall, user-group, or Origin-installation changes. When\n  Controlled Folder Access, an organization policy, cloud sync, or a read-only location blocks\n  writes, request access only to the affected folder or ask for an explicit alternate output folder.\n\n## Start with the beginner path\n\n1. Reject unsupported platforms before installing anything. Support the secure CLI\u002Fdependency layer only\n   on physical Windows 10\u002F11 x64 with 64-bit CPython 3.11 or 3.12; recommend 3.12 for new installs.\n   The locked Pillow version is 13.0.0; CPython 3.10 is no longer supported by this secure runtime.\n   Target Origin\u002FOriginPro 2021 and\n   later through external `originpro`; Origin 2020b and earlier are unsupported by this route.\n   Retain the complete CPython 3.10 + Origin 2024b \u002F 10.15 record as historical evidence only.\n   The current CPython 3.12.14 + Pillow 13.0.0 + Origin 2024b \u002F 10.15.132 live regression covers\n   six representative cases with smoke, complete artifacts, independent saved-project reopening,\n   data\u002Faxis\u002Fphysical-geometry readback, and re-exported PNG pixel identity checks. It does not\n   verify every route or Origin version; Python 3.11 has no live-verification claim. Require the\n   current workflow's smoke and artifact gates, and keep other 2021+ hosts capability-gated.\n   State plainly that macOS (Intel\u002FApple Silicon), Linux, WSL,\n   Wine\u002FCrossOver, Parallels, and other VMs are unsupported in V1. `doctor` cannot reliably detect\n   every VM, so ask the user to confirm a physical Windows host when that fact is unknown.\n2. Locate `editaplot.cmd` in the installed Skill directory; when working from a cloned repository,\n   use the repository-root `editaplot.cmd`. Use an absolute launcher path in commands. Do not make\n   beginners select a Python executable or invoke `scripts\u002Feditaplot.py` directly.\n3. This personal CC Switch distribution includes `editaplot.cmd` and a versioned\n   `.editaplot-local.json` that points to an audited, commit-named local runtime snapshot. Run\n   `py -3.12 scripts\u002Fvalidate_local_distribution.py`, then use the installed launcher; do not run upstream\n   `setup` against `.cc-switch`, `.codex`, or `.claude`, because CC Switch owns those runtime\n   directories. If validation reports a stale or missing engine, repair the source distribution and\n   resync it through CC Switch instead of patching a runtime copy. Read `references\u002Fruntime.md` for\n   discovery, update, and command details.\n4. Reuse an existing compatible Python 3.11 or 3.12. Prepare the new locked environment through the authorized CC Switch source-distribution workflow; retain the old environment for recovery, but never select it automatically after a failed upgrade. If none exists, explain in Chinese that installing Python\n   is a system-level change. Run `winget show` first and explain the exact publisher, source, and\n   agreements. Only explicit user confirmation permits a later non-interactive installation of\n   `Python.Python.3.12` with user scope and x64 architecture. If winget is unavailable, provide the\n   official python.org Windows installation instructions and wait for the user; never use an\n   untrusted mirror or silently install Python.\n5. Run `editaplot.cmd doctor` for each new workflow. Allow `doctor --repair` only for the reported\n   project-local Python dependency repair. Keep all Python packages in `.editaplot-venv`. Treat\n   Origin as a locally installed user-managed application; never install or modify it during repair.\n6. Run `editaplot.cmd start \u003Cdata-file>` for a new table. Add `--intent \"\u003Cuser intent>\"` when the\n   user states a goal. Treat its inspection and recommendation payload as internal working state.\n   Use the original local source path exposed by the attachment. If the host provides only a\n   temporary copied attachment and the original folder cannot be recovered, ask once for the intended\n   local source\u002Foutput folder before rendering; never guess an unrelated workspace destination.\n7. After selecting a candidate template, run `editaplot.cmd understand \u003Cdata-file>\n   --template-id \u003Cid>` with the same confirmed mapping that will be used for planning. Group its\n   result into an internal checklist: data type; columns to draw; columns used only for support or\n   validation; columns retained but not drawn; proposed figure elements; and calculations that\n   will **not** be performed. Every source column must appear exactly once. If any item is\n   `uncertain`, first check the supplied material; ask only about meaning still unresolved, then run\n   `understand` again. Do not confirm or plan unresolved items. Tell the user only what will be drawn\n   and any material question, unless they request the full column breakdown.\n8. Tell a beginner only: what was recognized, the best one to three chart choices, why they fit,\n   and the smallest scientific decision still required. Do not dump an\n   `inspect → recommend → understand → plan` pipeline or raw JSON unless they ask for technical\n   detail.\n9. Reuse the scientific purpose and element choices already explicit in the request. Ask only for unresolved scientific choices.\n   Freeze the exact `proposal_hash`, approved derived-item IDs, and resolved ambiguity choices in\n   `--semantic-confirmation-json`. Never reuse a confirmation after the source, mapping, purpose, or\n   proposal hash changes: regenerate the proposal and confirmation payload after revalidation.\n   A changed hash alone does not require another user answer. Reuse the existing decision when\n   the current request covers the updated data and its purpose, column roles, units, and authorized\n   transformations still match; ask only about an unresolved change in meaning. Internal\n   `awaiting_*_confirmation` states request a valid payload, not a mandatory extra conversation turn.\n   When confidence is low, candidate margins are small, roles or units are\n   ambiguous, or a new, unauthorized meaning-changing display transformation is proposed, ask only the additional focused questions\n   needed.\n10. If the user supplies a reference figure, first run `reference-inspect`. Codex may then describe\n   only its panel\u002Fmark\u002Fencoding\u002Flayout\u002Fstyle grammar in the strict ReferenceFigureSpec JSON and run\n   `reference-review`; the runtime performs no OCR or model inference. Show the adopted and rejected\n   features, bind every essential mark to confirmed renderable user data, and record the existing\n   exact request as authorization when it covers the selected features; ask only for new decisions. Never copy reference values, labels, fits, phase assignments, author\n   text, logos, watermarks, or the bitmap into the Origin project. Prefer verified\n   `template_adaptation`; keep `controlled_composition` blocked until that exact composition has\n   passed the full Origin evidence gate. A reference cannot add missing evidence or change the\n   confirmed scientific element list.\n   Reuse explicit palette, reference and style choices. With no preference, use the verified template\n   default; ask only for a material visual or scientific tradeoff. Record the actual request in the\n   existing confirmation interface, never invent a later user answer. Mark supported requests as\n   `applied`, `retained_template_default` or `rejected` and verify object readback.\n\n11. When the user needs palette options, run `editaplot.cmd palettes`, show\n   `assets\u002Fpalettes\u002Fpalette-selector-public.zh-CN.png`, and recommend no more than two compatible\n   `palette_id` values. Read `references\u002Fpalettes.md` before freezing one.\n12. Internally freeze the confirmed choice with `editaplot.cmd plan`; never hand-edit a plan or write\n   a decision back to the source file. For an exact XPS request, write the confirmed values to a\n   separate JSON object and pass its path with `--visual-style-json`. The supported exact fields are\n   `series_colors`, `line_width_pt`, `fill_transparency_percent`, `page_size_cm`, `legend_visible`,\n   `legend_position`, and `legend_frame`. Invalid explicit fields or values must fail fast and be\n   corrected with the user; never silently discard them or fall back to a reference\u002Fdefault style.\n   The render command copies this approved plan into the final output folder as `render-plan.json`.\n13. Treat Origin readiness as technical state only. Doctor performs read-only discovery of\n    `Origin.Application`, `Origin.ApplicationSI`, installed candidates, Python, `originpro`, and\n    `OriginExt`; it never launches Origin and `ready_for_render` never means a live connection\n    succeeded. If the default launch registration is present, proceed to the real pre-render smoke\n    without asking the user to open Origin or confirm it again. Keep beginner output to one to three\n    plain-language sentences; leave CLSIDs, registry views, candidates, and stages in JSON.\n14. Run `editaplot.cmd origin-smoke --output-dir \u003Cunique-smoke-directory>` with\n    `launch_isolated`: start and own a dedicated Origin instance, perform the live smoke and version\n    handshake, then apply the template capability decision. This command is mandatory after planning\n    and before formal rendering. `attach_existing` is an explicit advanced mode only; never reset,\n    overwrite, or close a user-owned project, and detach instead of exiting. Report failures by\n    technical stage and next step without speculation.\n    Never use mouse automation or provide application patches or bypass instructions. The runtime\n    must attempt to clean a partial EditaPlot-owned activation and may try one fresh isolated\n    instance for a retryable startup code only if cleanup succeeds. Cleanup failure returns\n    `origin_activation_cleanup_failed` and stops. It must then wait with `sec -poc 30` and confirm\n    `run.isOCready()` before reading the version or creating a project. Never loop, switch to\n    `ApplicationSI`, edit DCOM\u002Fregistry permissions, or tell a beginner to run the whole workflow as\n    administrator. After the automatic attempt is exhausted, request approval for at most one retry\n    in the same active Windows-user context and use a fresh empty sibling smoke directory so the\n    first report remains intact. `origin_com_class_not_registered` and\n    `origin_com_activation_access_denied` stop without automatic retry. Do not force-terminate a\n    Python worker merely because it has run for a long time: it may own a hidden Origin instance.\n    Preserve diagnostics and report the last progress stage before proposing any user-controlled\n    cancellation.\n15. Only after that smoke passes, render an allowed template route with\n    `editaplot.cmd render \u003Cplan>`. Keep an EditaPlot-owned Origin instance open after success unless\n    the user requests otherwise. By default, let the runtime create a direct sibling of the source\n    file named `\u003Csource_stem>_EditaPlot_YYYYMMDD_HHMMSS`; keep all formal artifacts in that folder.\n    Do not redirect ordinary runs to the repository, Skill directory, current working directory, or\n    a shared global output folder. Use `--output-dir` for an established task output folder or a\n    location explicitly requested by the user.\n16. Run `editaplot.cmd verify \u003Coutput-directory>` against the selected output folder and perform\n    agent visual QA. If smoke or render fails, a Python preview or standalone PNG\u002FPDF\u002FSVG is only\n    a preview and must not be presented as completed Origin work. Formal success requires the\n    editable OPJU, PNG, PDF, TIF, object readback, and agent visual QA together.\n\nBefore any render, read `references\u002Forigin-safety.md`, `references\u002Ffigure-contract.md`, and\n`references\u002Fverification.md`. For a new table or chart decision, read\n`references\u002Fdata-contracts.md`, `references\u002Fchart-selection.md`, and\n`references\u002Fsemantic-understanding.md`. When a reference image is supplied, also read\n`references\u002Freference-figures.md`.\n\n## Collaborate on a user-edited OPJU\n\nUse this route when the user wants to make final Origin edits personally and then asks for a\nreview. V1 covers data plots and already verified Origin templates; system diagrams and mechanism\nfigures remain in PPT\u002FSVG workflows.\n\n1. Run `editaplot.cmd prepare-opju-review \u003Cresult.opju>`. This creates a new sibling\n   `\u003Csource_stem>_OriginReview` workspace (or an explicitly requested new directory) containing\n   `figure_initial.opju`, `figure_edit.opju`, and `review-workspace.json`. It only copies and never\n   replaces the source, an existing workspace, or either copy. Treat `figure_initial.opju` as the\n   immutable baseline and `figure_edit.opju` as the only file the user edits and saves in Origin.\n2. Ask the user to edit and save `figure_edit.opju` directly in Origin. Do not require `v2\u002Fv3`\n   copies. Do not run the review command until the user explicitly says the file is saved and ready\n   for review.\n3. After that confirmation, run\n   `editaplot.cmd review-opju \u003Cworkspace>\\figure_edit.opju`. The command creates a unique\n   `review\u002F\u003Ctimestamp>\u002F` directory, copies `figure_edit.opju` to a snapshot, and checks SHA-256\n   before and after copying and after the isolated review. A changed, locked, or baseline-tampered\n   file stops with a structured error and never opens or overwrites the user file.\n4. The worker uses the fixed runtime's `OriginSession` with `new_isolated`, opens only the snapshot\n   using `readonly=True`, enumerates every Graph Page, and exports PNG\u002FPDF\u002FTIF below that review\n   directory. It must never attach to or save the user's Origin project; the session may initialize\n   its own empty EditaPlot-owned instance, and it closes only that instance.\n5. Briefly report the reviewed figures, usable export paths, and any unresolved issue affecting use;\n   keep `review-report.json` as a detailed record and link it only when useful or requested. It records snapshot and baseline\n   integrity, Origin version and instance ownership, Graph Page\u002Fobject inventory, and export hashes.\n   Export success means only “the snapshot can be read back and viewed”; inspect PNG\u002FTIF for\n   scientific meaning and visual quality and keep `human_visual_qa.status=pending` until the agent actually inspects the images.\n   V1 never writes changes back into the user's OPJU.\n\n## Keep scientific decisions with the user\n\n- Treat the original data file as immutable. Never overwrite it, fill missing source columns, or\n  invent measurements. Permit helper columns only in memory or the editable Origin project.\n- Classify every source column before planning as primary render, secondary render, support-only,\n  retain-not-render, or uncertain. Support-only and retained columns cannot become visible through\n  a reference image. An unknown numeric column is a question, not another automatic curve.\n- Distinguish scientific analysis from display transformation. Never silently normalize, smooth,\n  fit, remove outliers, calculate error bars, identify phases, or infer material peaks.\n- For GSAS\u002FGSAS-II Rietveld data, distinguish Observed, Calculated, optional Background, supplied\n  Difference, explicit Phase positions, and non-rendering control\u002Fdiagnostic columns. Preserve an\n  upstream Publication `Diff` exactly; never apply a second display offset.\n- For XPS, keep cosmetic preferences separate from the scientific contract. A user may explicitly\n  request exact series colors, physical line widths, fill transparency, a safe page\u002Faspect ratio,\n  and legend show\u002Fhide, borderless, or position choices. Apply only fields supported and read back\n  by the selected verified XPS renderer; otherwise retain the default or reject the field visibly.\n  Neither a user style request nor a reference image may change source values or column roles, the\n  high-to-low binding-energy axis contract, component identity, residual disposition, or the\n  verified single-region `set_fill_area(..., type=9)` \u002F `-pfm 3` fill implementation.\n- For FLIM, use `flim_phasor` for upstream G\u002FS coordinates with optional groups. Keep every point,\n  including points outside the reference semicircle. Confirm the mathematical semicircle helper;\n  preserve equal physical G\u002FS scales. Do not infer lifetimes, fit decays, unmix or cluster.\n- For supplied event times, use `event_raster` with explicit Time\u002Funit, Trial and Condition. Confirm\n  source-order trial ordinals; blank Time entries declare inventory, not events at zero. Preserve\n  every event and empty lane, and retain zero-event conditions in the borderless outside legend.\n  Never detect events, calculate PSTH\u002Frates or reorder trials by response. Read the event-table\n  section of `references\u002Fdata-contracts.md`; only condition-compatible palettes may replace the default.\n- For SHAP, accept only externally precomputed per-sample contributions. Never train a model or\n  invoke SHAP. Mean |SHAP| and optional group percentages may only summarize those supplied rows\n  with the allow-listed formulas recorded in the semantic proposal and explicitly approved; never\n  invent contributions or silently reorder features.\n- When the user requests separate importance bars, a beeswarm and a two-level contribution ring,\n  select `shap_dashboard`. Require explicit Feature Group, confirm feature\u002Fgroup percentages using\n  one total Mean |SHAP| denominator, and offer blue–white–wine, purple–green–yellow, or red–yellow–blue.\n  Freeze the selection as mapping plot_mode `dashboard_blue_red`, `dashboard_viridis`, or\n  `dashboard_red_blue`; do not substitute a general categorical palette for this continuous scale.\n- Confirm unknown units, error semantics, percentage denominators, meaningful order, dual axes,\n  and any other choice that can change the claim.\n- Recommend from the scientific question and data structure, not aesthetics alone. Refuse a\n  misleading chart even when technically renderable.\n- Keep template route status (`verified`, `experimental`, or `unsupported`) separate from current\n  host compatibility (`verified`, `compatible_unverified`, or `blocked`). Never relabel a\n  `compatible_unverified` Origin version as verified; continue only when its smoke succeeds and the\n  selected template's required capabilities are available.\n- Reject decorative 3D. Require a scientifically meaningful third axis; keep a new 3D route\n  experimental until Z-axis, camera, OpenGL type, source mapping, four exports, editable OPJU,\n  readback, and visual QA pass.\n- 对于已验证的 `density_ridgeline3d`，我只接受 2–6 个真实带单位条件的 mixed-wide\n  六角色表：上游提供同语义同单位的实线\u002F虚线预计算密度，并为每组提供恰好一个 `Focal X`。\n  焦点固定为 Z=0 基线 locator；不要计算 KDE、峰值、阈值、交点或焦点。当前主机还必须先通过\n  实时 smoke 与 `OPEN_GL_3D` 能力检查，不能只凭模板已验证就跳过主机门禁。\n- Do not send selected files to any additional network service or include them in public artifacts.\n  A file explicitly provided through Codex remains subject to the user's Codex account,\n  organization, and retention policies; do not claim the Skill can override those policies.\n- Before inspecting medical data or reference images, require the user to confirm that the material\n  follows their institution's rules, is deidentified, and has been checked for burned-in text.\n- Treat `panel-plan` as a deidentification-aware layout and evidence gate, not an OCR, PHI detector,\n  medical image editor, or merged editable Origin project. Preserve every verified subproject.\n\n## Apply the publication-informed contract\n\nFor supplied time–frequency tables, choose `time_frequency` only when time, frequency and value\ncolumns explicitly state their units and form a complete uniform grid. Preserve both coordinate\ndirections and every cell; use the numeric physical-axis route, not categorical `heatmap`.\nThe route shows sequential value magnitude with a detached scale; it performs no FFT\u002FSTFT\u002Fwavelet,\nfiltering, interpolation, baseline or dB conversion. A raw EEG voltage table needs upstream\nprocessing, not an invented frequency column. Read `references\u002Fdata-contracts.md` for the input\nand unit contract; confirm any extra columns instead of silently drawing them.\n\nFor multichannel EEG or fluorescence traces already supplied as a table, choose `signal_stack`.\nRequire explicit Time\u002FChannel\u002FAmplitude roles and units, keep every observation and the channel\norder, and confirm the additive display offsets and common unit ruler. Channel\u002FROI identities\nremain text, including leading zeros. Do not filter, align, center, normalize, infer events or\ndecode acquisition files. Direct names and the ruler live outside the true time-axis data range.\n\n- Make every chart defend one explicit conclusion or evidence role.\n- Use a white background, Arial, restrained color families, clear hierarchy, and no rainbow palette,\n  decorative 3D, or unjustified grid.\n- Derive physical Origin dimensions from chart type, data density, series count, and label length.\n  Keep fixed size only for a profile that explicitly requires it, such as legacy fixed C 1s.\n- Convert documented point, line-width, and page-size units correctly. Never copy small journal-page\n  font values directly into Origin API fields; read back the resulting axis and text objects.\n- Keep each condition's color consistent across related panels. Freeze palette IDs and exact HEX\n  values, allowed modes, safe category count, and accessibility constraints into the plan.\n- Let an explicit user style request outrank a style token inferred from a reference image. Color,\n  line-width, transparency, page\u002Faspect, and legend requests are still capability-gated and must be\n  classified as applied, retained default, or rejected before rendering.\n- Do not let a reference image or unverified cosmetic preference silently redefine semantic color\n  mappings for XPS components, signed effects, heatmaps, diagnostic lines, confusion matrices, or\n  similar evidence. An explicit replacement is allowed only through that route's independently\n  verified override with exact series mapping and readback.\n- Give every medical panel one distinct evidence role. Freeze a shared condition-to-color map before\n  composing quantitative panels; require explicit semantic confirmation for a shared legend.\n- Prefer editable labels and Origin objects. A Python preview or embedded bitmap is not an Origin\n  deliverable.\n- Call the result “publication-informed,” never “Nature compliant” or journal-approved.\n\n## Report the result in plain language\n\nDefault to a short Chinese result: what was drawn, links to the editable figure and useful exports,\nand any unresolved issue that affects use. If required validation is incomplete, say which part\nremains unverified. Do not list every alternative, confidence score, internal hash, plan or check file\nunless it helps the user decide or they request details. Keep data roles, authorized transformations,\nthe render plan, provenance and full validation records in the existing workflow's process files.\nConcise delivery does not remove any required artifact, object readback or agent visual QA.\n\n## Load detailed references only as needed\n\n- `references\u002Fruntime.md`: launcher, setup, Python discovery, CLI commands, and artifacts.\n- `references\u002Fchart-selection.md`: chart families, ranking rules, and support levels.\n- `references\u002Fdata-contracts.md`: accepted layouts, column semantics, and repair guidance.\n- `references\u002Fsemantic-understanding.md`: per-column use, element checklist, derived-data lineage,\n  and the hash-bound confirmation gate.\n- `references\u002Freference-figures.md`: safe reference grammar, bindings, adaptation limits, and\n  recording of the existing exact authorization.\n- `references\u002Ffigure-contract.md`: evidence logic, visual hierarchy, typography, and color rules.\n- `references\u002Forigin-safety.md`: local Automation and verified-API guardrails.\n- `references\u002Fverification.md`: mandatory artifacts, readback, and visual QA.\n- `references\u002Fshowcase.md`: neutral demonstration data and gallery policy.\n- `references\u002Fpalettes.md`: Chinese palette selector, compatibility, and accessibility limits.\n","# EditaPlot\n\n## 与论文图件规范的分工\n\n单独提出科研数据绘图请求时，在当前 Origin 路线能满足要求的前提下优先使用本技能，交付用户可继续手工微调的原生 OPJU；无需用户明确说 Origin。论文结果图由 `paper-figure-review` 确定技术含义、坐标单位和投稿排版要求，本技能负责导入、制作、保存和导出，再按该要求检查实际输出。\n\n论文任务已确定的字体、字号、尺寸等要求优先于下文及参考资料的 Origin 模板默认值；当前路线不能实现或回读验证时，报告缺口。跨技能继承已确定的任务输出目录，用现有输出参数传入，无需再次询问目录。\n\n科研示意图走 PPT、SVG 等设计路线。仿真、实验程序运行时的自动出图保留现有实现；用户指定 Python\u002FMATLAB 或沿用既有图源时遵循指定，不自动迁移。仅审查时给意见，不生成修改版。\n\n两个技能复用同一任务的数据含义、样式和授权，不各问一轮确认。用户保存后的当前 OPJU 是后续读取依据，按现有副本审查流程导出，不覆盖其手工调整；这不代表已提供任意 OPJU 原位编辑接口。\n\n制作前核对必要坐标尺度、阈值线、标记、版面与导出格式是否被当前路线支持；仅模板名匹配或 Origin 可启动不算满足。能力不足时说明具体缺口和可行替代，不静默把对数轴换成线性轴，也不把不符合要求的试图标为完成。\n\nTurn a scientific question and a read-only table into an auditable, editable Origin figure. Keep\nthe beginner experience conversational; use the deterministic engine for inspection, planning,\nrendering, exporting, and readback.\n\n## Request only scoped Windows permissions\n\n- Read the configured versioned runtime, selected table, and optional local reference image. Read the\n  complete repository only when working from an explicitly selected clone or approved source update.\n- Write only to the configured runtime's `.editaplot-venv`, an approved source repository during a\n  separately authorized update, and the established task output folder (defaulting to the selected\n  source file's parent folder). Never patch `.cc-switch`, `.codex`, or `.claude` Skill copies directly.\n- Run the local launcher, PowerShell\u002FPython subprocesses, and an EditaPlot-owned Origin instance in\n  the same active interactive Windows user session.\n- Use network access only for repository download\u002Fupdate and locked dependency retrieval. Treat a\n  user-scope winget Python installation as a separate system change that still requires explicit\n  consent.\n- Do not request administrator rights, mouse control, whole-drive write access, cloud upload of\n  private inputs, or DCOM, registry, firewall, user-group, or Origin-installation changes. When\n  Controlled Folder Access, an organization policy, cloud sync, or a read-only location blocks\n  writes, request access only to the affected folder or ask for an explicit alternate output folder.\n\n## Start with the beginner path\n\n1. Reject unsupported platforms before installing anything. Support the secure CLI\u002Fdependency layer only\n   on physical Windows 10\u002F11 x64 with 64-bit CPython 3.11 or 3.12; recommend 3.12 for new installs.\n   The locked Pillow version is 13.0.0; CPython 3.10 is no longer supported by this secure runtime.\n   Target Origin\u002FOriginPro 2021 and\n   later through external `originpro`; Origin 2020b and earlier are unsupported by this route.\n   Retain the complete CPython 3.10 + Origin 2024b \u002F 10.15 record as historical evidence only.\n   The current CPython 3.12.14 + Pillow 13.0.0 + Origin 2024b \u002F 10.15.132 live regression covers\n   six representative cases with smoke, complete artifacts, independent saved-project reopening,\n   data\u002Faxis\u002Fphysical-geometry readback, and re-exported PNG pixel identity checks. It does not\n   verify every route or Origin version; Python 3.11 has no live-verification claim. Require the\n   current workflow's smoke and artifact gates, and keep other 2021+ hosts capability-gated.\n   State plainly that macOS (Intel\u002FApple Silicon), Linux, WSL,\n   Wine\u002FCrossOver, Parallels, and other VMs are unsupported in V1. `doctor` cannot reliably detect\n   every VM, so ask the user to confirm a physical Windows host when that fact is unknown.\n2. Locate `editaplot.cmd` in the installed Skill directory; when working from a cloned repository,\n   use the repository-root `editaplot.cmd`. Use an absolute launcher path in commands. Do not make\n   beginners select a Python executable or invoke `scripts\u002Feditaplot.py` directly.\n3. This personal CC Switch distribution includes `editaplot.cmd` and a versioned\n   `.editaplot-local.json` that points to an audited, commit-named local runtime snapshot. Run\n   `py -3.12 scripts\u002Fvalidate_local_distribution.py`, then use the installed launcher; do not run upstream\n   `setup` against `.cc-switch`, `.codex`, or `.claude`, because CC Switch owns those runtime\n   directories. If validation reports a stale or missing engine, repair the source distribution and\n   resync it through CC Switch instead of patching a runtime copy. Read `references\u002Fruntime.md` for\n   discovery, update, and command details.\n4. Reuse an existing compatible Python 3.11 or 3.12. Prepare the new locked environment through the authorized CC Switch source-distribution workflow; retain the old environment for recovery, but never select it automatically after a failed upgrade. If none exists, explain in Chinese that installing Python\n   is a system-level change. Run `winget show` first and explain the exact publisher, source, and\n   agreements. Only explicit user confirmation permits a later non-interactive installation of\n   `Python.Python.3.12` with user scope and x64 architecture. If winget is unavailable, provide the\n   official python.org Windows installation instructions and wait for the user; never use an\n   untrusted mirror or silently install Python.\n5. Run `editaplot.cmd doctor` for each new workflow. Allow `doctor --repair` only for the reported\n   project-local Python dependency repair. Keep all Python packages in `.editaplot-venv`. Treat\n   Origin as a locally installed user-managed application; never install or modify it during repair.\n6. Run `editaplot.cmd start \u003Cdata-file>` for a new table. Add `--intent \"\u003Cuser intent>\"` when the\n   user states a goal. Treat its inspection and recommendation payload as internal working state.\n   Use the original local source path exposed by the attachment. If the host provides only a\n   temporary copied attachment and the original folder cannot be recovered, ask once for the intended\n   local source\u002Foutput folder before rendering; never guess an unrelated workspace destination.\n7. After selecting a candidate template, run `editaplot.cmd understand \u003Cdata-file>\n   --template-id \u003Cid>` with the same confirmed mapping that will be used for planning. Group its\n   result into an internal checklist: data type; columns to draw; columns used only for support or\n   validation; columns retained but not drawn; proposed figure elements; and calculations that\n   will **not** be performed. Every source column must appear exactly once. If any item is\n   `uncertain`, first check the supplied material; ask only about meaning still unresolved, then run\n   `understand` again. Do not confirm or plan unresolved items. Tell the user only what will be drawn\n   and any material question, unless they request the full column breakdown.\n8. Tell a beginner only: what was recognized, the best one to three chart choices, why they fit,\n   and the smallest scientific decision still required. Do not dump an\n   `inspect → recommend → understand → plan` pipeline or raw JSON unless they ask for technical\n   detail.\n9. Reuse the scientific purpose and element choices already explicit in the request. Ask only for unresolved scientific choices.\n   Freeze the exact `proposal_hash`, approved derived-item IDs, and resolved ambiguity choices in\n   `--semantic-confirmation-json`. Never reuse a confirmation after the source, mapping, purpose, or\n   proposal hash changes: regenerate the proposal and confirmation payload after revalidation.\n   A changed hash alone does not require another user answer. Reuse the existing decision when\n   the current request covers the updated data and its purpose, column roles, units, and authorized\n   transformations still match; ask only about an unresolved change in meaning. Internal\n   `awaiting_*_confirmation` states request a valid payload, not a mandatory extra conversation turn.\n   When confidence is low, candidate margins are small, roles or units are\n   ambiguous, or a new, unauthorized meaning-changing display transformation is proposed, ask only the additional focused questions\n   needed.\n10. If the user supplies a reference figure, first run `reference-inspect`. Codex may then describe\n   only its panel\u002Fmark\u002Fencoding\u002Flayout\u002Fstyle grammar in the strict ReferenceFigureSpec JSON and run\n   `reference-review`; the runtime performs no OCR or model inference. Show the adopted and rejected\n   features, bind every essential mark to confirmed renderable user data, and record the existing\n   exact request as authorization when it covers the selected features; ask only for new decisions. Never copy reference values, labels, fits, phase assignments, author\n   text, logos, watermarks, or the bitmap into the Origin project. Prefer verified\n   `template_adaptation`; keep `controlled_composition` blocked until that exact composition has\n   passed the full Origin evidence gate. A reference cannot add missing evidence or change the\n   confirmed scientific element list.\n   Reuse explicit palette, reference and style choices. With no preference, use the verified template\n   default; ask only for a material visual or scientific tradeoff. Record the actual request in the\n   existing confirmation interface, never invent a later user answer. Mark supported requests as\n   `applied`, `retained_template_default` or `rejected` and verify object readback.\n\n11. When the user needs palette options, run `editaplot.cmd palettes`, show\n   `assets\u002Fpalettes\u002Fpalette-selector-public.zh-CN.png`, and recommend no more than two compatible\n   `palette_id` values. Read `references\u002Fpalettes.md` before freezing one.\n12. Internally freeze the confirmed choice with `editaplot.cmd plan`; never hand-edit a plan or write\n   a decision back to the source file. For an exact XPS request, write the confirmed values to a\n   separate JSON object and pass its path with `--visual-style-json`. The supported exact fields are\n   `series_colors`, `line_width_pt`, `fill_transparency_percent`, `page_size_cm`, `legend_visible`,\n   `legend_position`, and `legend_frame`. Invalid explicit fields or values must fail fast and be\n   corrected with the user; never silently discard them or fall back to a reference\u002Fdefault style.\n   The render command copies this approved plan into the final output folder as `render-plan.json`.\n13. Treat Origin readiness as technical state only. Doctor performs read-only discovery of\n    `Origin.Application`, `Origin.ApplicationSI`, installed candidates, Python, `originpro`, and\n    `OriginExt`; it never launches Origin and `ready_for_render` never means a live connection\n    succeeded. If the default launch registration is present, proceed to the real pre-render smoke\n    without asking the user to open Origin or confirm it again. Keep beginner output to one to three\n    plain-language sentences; leave CLSIDs, registry views, candidates, and stages in JSON.\n14. Run `editaplot.cmd origin-smoke --output-dir \u003Cunique-smoke-directory>` with\n    `launch_isolated`: start and own a dedicated Origin instance, perform the live smoke and version\n    handshake, then apply the template capability decision. This command is mandatory after planning\n    and before formal rendering. `attach_existing` is an explicit advanced mode only; never reset,\n    overwrite, or close a user-owned project, and detach instead of exiting. Report failures by\n    technical stage and next step without speculation.\n    Never use mouse automation or provide application patches or bypass instructions. The runtime\n    must attempt to clean a partial EditaPlot-owned activation and may try one fresh isolated\n    instance for a retryable startup code only if cleanup succeeds. Cleanup failure returns\n    `origin_activation_cleanup_failed` and stops. It must then wait with `sec -poc 30` and confirm\n    `run.isOCready()` before reading the version or creating a project. Never loop, switch to\n    `ApplicationSI`, edit DCOM\u002Fregistry permissions, or tell a beginner to run the whole workflow as\n    administrator. After the automatic attempt is exhausted, request approval for at most one retry\n    in the same active Windows-user context and use a fresh empty sibling smoke directory so the\n    first report remains intact. `origin_com_class_not_registered` and\n    `origin_com_activation_access_denied` stop without automatic retry. Do not force-terminate a\n    Python worker merely because it has run for a long time: it may own a hidden Origin instance.\n    Preserve diagnostics and report the last progress stage before proposing any user-controlled\n    cancellation.\n15. Only after that smoke passes, render an allowed template route with\n    `editaplot.cmd render \u003Cplan>`. Keep an EditaPlot-owned Origin instance open after success unless\n    the user requests otherwise. By default, let the runtime create a direct sibling of the source\n    file named `\u003Csource_stem>_EditaPlot_YYYYMMDD_HHMMSS`; keep all formal artifacts in that folder.\n    Do not redirect ordinary runs to the repository, Skill directory, current working directory, or\n    a shared global output folder. Use `--output-dir` for an established task output folder or a\n    location explicitly requested by the user.\n16. Run `editaplot.cmd verify \u003Coutput-directory>` against the selected output folder and perform\n    agent visual QA. If smoke or render fails, a Python preview or standalone PNG\u002FPDF\u002FSVG is only\n    a preview and must not be presented as completed Origin work. Formal success requires the\n    editable OPJU, PNG, PDF, TIF, object readback, and agent visual QA together.\n\nBefore any render, read `references\u002Forigin-safety.md`, `references\u002Ffigure-contract.md`, and\n`references\u002Fverification.md`. For a new table or chart decision, read\n`references\u002Fdata-contracts.md`, `references\u002Fchart-selection.md`, and\n`references\u002Fsemantic-understanding.md`. When a reference image is supplied, also read\n`references\u002Freference-figures.md`.\n\n## Collaborate on a user-edited OPJU\n\nUse this route when the user wants to make final Origin edits personally and then asks for a\nreview. V1 covers data plots and already verified Origin templates; system diagrams and mechanism\nfigures remain in PPT\u002FSVG workflows.\n\n1. Run `editaplot.cmd prepare-opju-review \u003Cresult.opju>`. This creates a new sibling\n   `\u003Csource_stem>_OriginReview` workspace (or an explicitly requested new directory) containing\n   `figure_initial.opju`, `figure_edit.opju`, and `review-workspace.json`. It only copies and never\n   replaces the source, an existing workspace, or either copy. Treat `figure_initial.opju` as the\n   immutable baseline and `figure_edit.opju` as the only file the user edits and saves in Origin.\n2. Ask the user to edit and save `figure_edit.opju` directly in Origin. Do not require `v2\u002Fv3`\n   copies. Do not run the review command until the user explicitly says the file is saved and ready\n   for review.\n3. After that confirmation, run\n   `editaplot.cmd review-opju \u003Cworkspace>\\figure_edit.opju`. The command creates a unique\n   `review\u002F\u003Ctimestamp>\u002F` directory, copies `figure_edit.opju` to a snapshot, and checks SHA-256\n   before and after copying and after the isolated review. A changed, locked, or baseline-tampered\n   file stops with a structured error and never opens or overwrites the user file.\n4. The worker uses the fixed runtime's `OriginSession` with `new_isolated`, opens only the snapshot\n   using `readonly=True`, enumerates every Graph Page, and exports PNG\u002FPDF\u002FTIF below that review\n   directory. It must never attach to or save the user's Origin project; the session may initialize\n   its own empty EditaPlot-owned instance, and it closes only that instance.\n5. Briefly report the reviewed figures, usable export paths, and any unresolved issue affecting use;\n   keep `review-report.json` as a detailed record and link it only when useful or requested. It records snapshot and baseline\n   integrity, Origin version and instance ownership, Graph Page\u002Fobject inventory, and export hashes.\n   Export success means only “the snapshot can be read back and viewed”; inspect PNG\u002FTIF for\n   scientific meaning and visual quality and keep `human_visual_qa.status=pending` until the agent actually inspects the images.\n   V1 never writes changes back into the user's OPJU.\n\n## Keep scientific decisions with the user\n\n- Treat the original data file as immutable. Never overwrite it, fill missing source columns, or\n  invent measurements. Permit helper columns only in memory or the editable Origin project.\n- Classify every source column before planning as primary render, secondary render, support-only,\n  retain-not-render, or uncertain. Support-only and retained columns cannot become visible through\n  a reference image. An unknown numeric column is a question, not another automatic curve.\n- Distinguish scientific analysis from display transformation. Never silently normalize, smooth,\n  fit, remove outliers, calculate error bars, identify phases, or infer material peaks.\n- For GSAS\u002FGSAS-II Rietveld data, distinguish Observed, Calculated, optional Background, supplied\n  Difference, explicit Phase positions, and non-rendering control\u002Fdiagnostic columns. Preserve an\n  upstream Publication `Diff` exactly; never apply a second display offset.\n- For XPS, keep cosmetic preferences separate from the scientific contract. A user may explicitly\n  request exact series colors, physical line widths, fill transparency, a safe page\u002Faspect ratio,\n  and legend show\u002Fhide, borderless, or position choices. Apply only fields supported and read back\n  by the selected verified XPS renderer; otherwise retain the default or reject the field visibly.\n  Neither a user style request nor a reference image may change source values or column roles, the\n  high-to-low binding-energy axis contract, component identity, residual disposition, or the\n  verified single-region `set_fill_area(..., type=9)` \u002F `-pfm 3` fill implementation.\n- For FLIM, use `flim_phasor` for upstream G\u002FS coordinates with optional groups. Keep every point,\n  including points outside the reference semicircle. Confirm the mathematical semicircle helper;\n  preserve equal physical G\u002FS scales. Do not infer lifetimes, fit decays, unmix or cluster.\n- For supplied event times, use `event_raster` with explicit Time\u002Funit, Trial and Condition. Confirm\n  source-order trial ordinals; blank Time entries declare inventory, not events at zero. Preserve\n  every event and empty lane, and retain zero-event conditions in the borderless outside legend.\n  Never detect events, calculate PSTH\u002Frates or reorder trials by response. Read the event-table\n  section of `references\u002Fdata-contracts.md`; only condition-compatible palettes may replace the default.\n- For SHAP, accept only externally precomputed per-sample contributions. Never train a model or\n  invoke SHAP. Mean |SHAP| and optional group percentages may only summarize those supplied rows\n  with the allow-listed formulas recorded in the semantic proposal and explicitly approved; never\n  invent contributions or silently reorder features.\n- When the user requests separate importance bars, a beeswarm and a two-level contribution ring,\n  select `shap_dashboard`. Require explicit Feature Group, confirm feature\u002Fgroup percentages using\n  one total Mean |SHAP| denominator, and offer blue–white–wine, purple–green–yellow, or red–yellow–blue.\n  Freeze the selection as mapping plot_mode `dashboard_blue_red`, `dashboard_viridis`, or\n  `dashboard_red_blue`; do not substitute a general categorical palette for this continuous scale.\n- Confirm unknown units, error semantics, percentage denominators, meaningful order, dual axes,\n  and any other choice that can change the claim.\n- Recommend from the scientific question and data structure, not aesthetics alone. Refuse a\n  misleading chart even when technically renderable.\n- Keep template route status (`verified`, `experimental`, or `unsupported`) separate from current\n  host compatibility (`verified`, `compatible_unverified`, or `blocked`). Never relabel a\n  `compatible_unverified` Origin version as verified; continue only when its smoke succeeds and the\n  selected template's required capabilities are available.\n- Reject decorative 3D. Require a scientifically meaningful third axis; keep a new 3D route\n  experimental until Z-axis, camera, OpenGL type, source mapping, four exports, editable OPJU,\n  readback, and visual QA pass.\n- 对于已验证的 `density_ridgeline3d`，我只接受 2–6 个真实带单位条件的 mixed-wide\n  六角色表：上游提供同语义同单位的实线\u002F虚线预计算密度，并为每组提供恰好一个 `Focal X`。\n  焦点固定为 Z=0 基线 locator；不要计算 KDE、峰值、阈值、交点或焦点。当前主机还必须先通过\n  实时 smoke 与 `OPEN_GL_3D` 能力检查，不能只凭模板已验证就跳过主机门禁。\n- Do not send selected files to any additional network service or include them in public artifacts.\n  A file explicitly provided through Codex remains subject to the user's Codex account,\n  organization, and retention policies; do not claim the Skill can override those policies.\n- Before inspecting medical data or reference images, require the user to confirm that the material\n  follows their institution's rules, is deidentified, and has been checked for burned-in text.\n- Treat `panel-plan` as a deidentification-aware layout and evidence gate, not an OCR, PHI detector,\n  medical image editor, or merged editable Origin project. Preserve every verified subproject.\n\n## Apply the publication-informed contract\n\nFor supplied time–frequency tables, choose `time_frequency` only when time, frequency and value\ncolumns explicitly state their units and form a complete uniform grid. Preserve both coordinate\ndirections and every cell; use the numeric physical-axis route, not categorical `heatmap`.\nThe route shows sequential value magnitude with a detached scale; it performs no FFT\u002FSTFT\u002Fwavelet,\nfiltering, interpolation, baseline or dB conversion. A raw EEG voltage table needs upstream\nprocessing, not an invented frequency column. Read `references\u002Fdata-contracts.md` for the input\nand unit contract; confirm any extra columns instead of silently drawing them.\n\nFor multichannel EEG or fluorescence traces already supplied as a table, choose `signal_stack`.\nRequire explicit Time\u002FChannel\u002FAmplitude roles and units, keep every observation and the channel\norder, and confirm the additive display offsets and common unit ruler. Channel\u002FROI identities\nremain text, including leading zeros. Do not filter, align, center, normalize, infer events or\ndecode acquisition files. Direct names and the ruler live outside the true time-axis data range.\n\n- Make every chart defend one explicit conclusion or evidence role.\n- Use a white background, Arial, restrained color families, clear hierarchy, and no rainbow palette,\n  decorative 3D, or unjustified grid.\n- Derive physical Origin dimensions from chart type, data density, series count, and label length.\n  Keep fixed size only for a profile that explicitly requires it, such as legacy fixed C 1s.\n- Convert documented point, line-width, and page-size units correctly. Never copy small journal-page\n  font values directly into Origin API fields; read back the resulting axis and text objects.\n- Keep each condition's color consistent across related panels. Freeze palette IDs and exact HEX\n  values, allowed modes, safe category count, and accessibility constraints into the plan.\n- Let an explicit user style request outrank a style token inferred from a reference image. Color,\n  line-width, transparency, page\u002Faspect, and legend requests are still capability-gated and must be\n  classified as applied, retained default, or rejected before rendering.\n- Do not let a reference image or unverified cosmetic preference silently redefine semantic color\n  mappings for XPS components, signed effects, heatmaps, diagnostic lines, confusion matrices, or\n  similar evidence. An explicit replacement is allowed only through that route's independently\n  verified override with exact series mapping and readback.\n- Give every medical panel one distinct evidence role. Freeze a shared condition-to-color map before\n  composing quantitative panels; require explicit semantic confirmation for a shared legend.\n- Prefer editable labels and Origin objects. A Python preview or embedded bitmap is not an Origin\n  deliverable.\n- Call the result “publication-informed,” never “Nature compliant” or journal-approved.\n\n## Report the result in plain language\n\nDefault to a short Chinese result: what was drawn, links to the editable figure and useful exports,\nand any unresolved issue that affects use. If required validation is incomplete, say which part\nremains unverified. Do not list every alternative, confidence score, internal hash, plan or check file\nunless it helps the user decide or they request details. Keep data roles, authorized transformations,\nthe render plan, provenance and full validation records in the existing workflow's process files.\nConcise delivery does not remove any required artifact, object readback or agent visual QA.\n\n## Load detailed references only as needed\n\n- `references\u002Fruntime.md`: launcher, setup, Python discovery, CLI commands, and artifacts.\n- `references\u002Fchart-selection.md`: chart families, ranking rules, and support levels.\n- `references\u002Fdata-contracts.md`: accepted layouts, column semantics, and repair guidance.\n- `references\u002Fsemantic-understanding.md`: per-column use, element checklist, derived-data lineage,\n  and the hash-bound confirmation gate.\n- `references\u002Freference-figures.md`: safe reference grammar, bindings, adaptation limits, and\n  recording of the existing exact authorization.\n- `references\u002Ffigure-contract.md`: evidence logic, visual hierarchy, typography, and color rules.\n- `references\u002Forigin-safety.md`: local Automation and verified-API guardrails.\n- `references\u002Fverification.md`: mandatory artifacts, readback, and visual QA.\n- `references\u002Fshowcase.md`: neutral demonstration data and gallery policy.\n- `references\u002Fpalettes.md`: Chinese palette selector, compatibility, and accessibility limits.\n",{"name":5,"description":7},"main","35fe63ba8f8645d64cc3b963bb5d8b6aca5c7db5","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills","https:\u002F\u002Fgithub.com\u002FSanAntonio021","2026-11-14 03:01:39",{"source":8,"skillName":5,"commands":63,"prompt":73,"zipAvailable":74},[64,67,70],{"tool":65,"command":66},"npx","npx skills add SanAntonio021\u002Fagents-skills --skill editaplot",{"tool":68,"command":69},"bunx","bunx skills add SanAntonio021\u002Fagents-skills --skill editaplot",{"tool":71,"command":72},"pnpm","pnpm dlx skills add SanAntonio021\u002Fagents-skills --skill editaplot","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002FSKILL.md",true,[76,82,86,92,96,100,104,108,112,116,120,124,128,132,136,140,146,150,154,158,162,166,170,174,178,182,186,190,194,198,204,208,212,216,220,224,228,232],{"path":77,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":80,"sourceHtmlUrl":81},".editaplot-local.json","REFERENCE","参考文档",448,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002F.editaplot-local.json",{"path":83,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":84,"sourceHtmlUrl":85},"agents\u002Fopenai.yaml",791,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fagents\u002Fopenai.yaml",{"path":87,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":90,"sourceHtmlUrl":91},"assets\u002Fpalettes\u002Fcards\u002Famber_lavender.png","ASSET","资产文件",55442,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Famber_lavender.png",{"path":93,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":94,"sourceHtmlUrl":95},"assets\u002Fpalettes\u002Fcards\u002Fblue_coral.png",55700,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Fblue_coral.png",{"path":97,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":98,"sourceHtmlUrl":99},"assets\u002Fpalettes\u002Fcards\u002Fdeep_sea_gold.png",54927,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Fdeep_sea_gold.png",{"path":101,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":102,"sourceHtmlUrl":103},"assets\u002Fpalettes\u002Fcards\u002Fforest_amber.png",54831,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Fforest_amber.png",{"path":105,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":106,"sourceHtmlUrl":107},"assets\u002Fpalettes\u002Fcards\u002Fnavy_cyan_gold.png",53094,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Fnavy_cyan_gold.png",{"path":109,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":110,"sourceHtmlUrl":111},"assets\u002Fpalettes\u002Fcards\u002Fnavy_ember.png",55579,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Fnavy_ember.png",{"path":113,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":114,"sourceHtmlUrl":115},"assets\u002Fpalettes\u002Fcards\u002Focean_coral.png",56220,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Focean_coral.png",{"path":117,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":118,"sourceHtmlUrl":119},"assets\u002Fpalettes\u002Fcards\u002Fplum_rose.png",56933,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Fplum_rose.png",{"path":121,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":122,"sourceHtmlUrl":123},"assets\u002Fpalettes\u002Fcards\u002Fsky_terra.png",53238,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Fsky_terra.png",{"path":125,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":126,"sourceHtmlUrl":127},"assets\u002Fpalettes\u002Fcards\u002Fviolet_lime.png",53055,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fcards\u002Fviolet_lime.png",{"path":129,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":130,"sourceHtmlUrl":131},"assets\u002Fpalettes\u002Fpalette-catalog.json",9063,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fpalette-catalog.json",{"path":133,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":134,"sourceHtmlUrl":135},"assets\u002Fpalettes\u002Fpalette-selector-all.zh-CN.png",469526,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fpalette-selector-all.zh-CN.png",{"path":137,"resourceType":88,"resourceTypeDesc":89,"sizeBytes":138,"sourceHtmlUrl":139},"assets\u002Fpalettes\u002Fpalette-selector-public.zh-CN.png",387630,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fassets\u002Fpalettes\u002Fpalette-selector-public.zh-CN.png",{"path":141,"resourceType":142,"resourceTypeDesc":143,"sizeBytes":144,"sourceHtmlUrl":145},"editaplot.cmd","OTHER","其他",7405,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Feditaplot.cmd",{"path":147,"resourceType":147,"resourceTypeDesc":148,"sizeBytes":149,"sourceHtmlUrl":16},"LICENSE","许可证",11357,{"path":151,"resourceType":142,"resourceTypeDesc":143,"sizeBytes":152,"sourceHtmlUrl":153},"NOTICE",861,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002FNOTICE",{"path":155,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":156,"sourceHtmlUrl":157},"references\u002Fchart-selection.md",15341,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Fchart-selection.md",{"path":159,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":160,"sourceHtmlUrl":161},"references\u002Fdata-contracts.md",22880,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Fdata-contracts.md",{"path":163,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":164,"sourceHtmlUrl":165},"references\u002Ffigure-contract.md",6032,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Ffigure-contract.md",{"path":167,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":168,"sourceHtmlUrl":169},"references\u002Forigin-safety.md",11814,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Forigin-safety.md",{"path":171,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":172,"sourceHtmlUrl":173},"references\u002Fpalettes.md",2968,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Fpalettes.md",{"path":175,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":176,"sourceHtmlUrl":177},"references\u002Freference-figures.md",7897,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Freference-figures.md",{"path":179,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":180,"sourceHtmlUrl":181},"references\u002Fruntime.md",23800,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Fruntime.md",{"path":183,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":184,"sourceHtmlUrl":185},"references\u002Fsemantic-understanding.md",4290,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Fsemantic-understanding.md",{"path":187,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":188,"sourceHtmlUrl":189},"references\u002Fshowcase.md",4867,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Fshowcase.md",{"path":191,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":192,"sourceHtmlUrl":193},"references\u002Fupstream-sources.md",2013,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Fupstream-sources.md",{"path":195,"resourceType":78,"resourceTypeDesc":79,"sizeBytes":196,"sourceHtmlUrl":197},"references\u002Fverification.md",6023,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Freferences\u002Fverification.md",{"path":199,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":202,"sourceHtmlUrl":203},"scripts\u002Fbootstrap_editaplot.py","SCRIPT","脚本",45694,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fscripts\u002Fbootstrap_editaplot.py",{"path":205,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":206,"sourceHtmlUrl":207},"scripts\u002Feditaplot_core.py",168139,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fscripts\u002Feditaplot_core.py",{"path":209,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":210,"sourceHtmlUrl":211},"scripts\u002Feditaplot.py",23086,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fscripts\u002Feditaplot.py",{"path":213,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":214,"sourceHtmlUrl":215},"scripts\u002Flaunch_managed.ps1",1564,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fscripts\u002Flaunch_managed.ps1",{"path":217,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":218,"sourceHtmlUrl":219},"scripts\u002Fopju_review.py",38297,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fscripts\u002Fopju_review.py",{"path":221,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":222,"sourceHtmlUrl":223},"scripts\u002Frequirements-runtime.lock",629,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fscripts\u002Frequirements-runtime.lock",{"path":225,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":226,"sourceHtmlUrl":227},"scripts\u002Fvalidate_local_distribution.py",8905,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Fscripts\u002Fvalidate_local_distribution.py",{"path":229,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":230,"sourceHtmlUrl":231},"tests\u002Ftest_opju_review.py",10496,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Ftests\u002Ftest_opju_review.py",{"path":233,"resourceType":200,"resourceTypeDesc":201,"sizeBytes":234,"sourceHtmlUrl":235},"tests\u002Ftest_validate_local_distribution.py",5147,"https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Feditaplot\u002Ftests\u002Ftest_validate_local_distribution.py",[237,266,284,309,319,342],{"skillNo":238,"name":239,"nameZh":6,"description":240,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":241,"licenseFileUrl":242,"sourceHtmlUrl":243,"categories":244,"keywords":254,"publishedAt":52},"01M71F7NWCYRJG4C1WVFSDPS2H","writing-router","统一写作协作流程，并处理普通中文编辑。Use when 用户要撰写、重写、润色或审查项目书、技术方案、会前技术交流稿、系统说明、专利技术交底书、测试与结果分析、调研报告、会议纪要、中文或英文论文，先确定主稿、修改范围和协作方式，再结合对应专业技能完成正文、检查与写回；也直接处理界面文案、普通中文去 AI 味、删废话和无法直接归类的中文材料。投稿事务、停稿审查、文献检索和单纯文件排版由对应专门技能处理。","MIT","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fwriting-router\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fwriting-router\u002FSKILL.md",[245,248,251],{"categoryNo":246,"categoryName":247,"taxonomyType":22,"taxonomyTypeDesc":23},"01JBSEEDCAT000000000000008","测试与质量",{"categoryNo":249,"categoryName":250,"taxonomyType":22,"taxonomyTypeDesc":23},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":252,"categoryName":253,"taxonomyType":30,"taxonomyTypeDesc":31},"01JBSEEDCAT000000000000036","内容创作者",[255,256,257,258,259,260,261,262,263,264,265],"writing","writing router","router","文案","中文","testing","报告","文件","use","协作","when",{"skillNo":267,"name":268,"nameZh":6,"description":269,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":241,"licenseFileUrl":6,"sourceHtmlUrl":270,"categories":271,"keywords":272,"publishedAt":52},"01M71F7F7J66P4K9SKXR9JD6KD","web-access","统一处理联网任务的工具选择，包括搜索、网页读取、下载和浏览器操作。 普通检索直接使用可用搜索或读取工具；抓取受阻、正文不完整时主动转到浏览器继续。 已知依赖登录态、动态内容或页面交互时直接使用浏览器，无需用户再次点名。","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fweb-access\u002FSKILL.md",[],[273,274,275,276,277,278,279,280,281,282,283],"access","web","web access","edge","工具","搜索","github","Existing","交互","下载","https",{"skillNo":285,"name":286,"nameZh":6,"description":287,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":241,"licenseFileUrl":288,"sourceHtmlUrl":289,"categories":290,"keywords":294,"publishedAt":52},"01M71F7C4AP46V8XEM7XX0HFYW","visiomaster","在 Windows 上将已确认的文字讨论或参考图片绘制为可编辑 Visio 流程图、系统架构图和端口连接图，并打开已有 .vsdx 按页面和 Shape ID 局部修改原生模块、文字、分组及连线，保留其他对象。用户要求画 Visio、生成 .vsdx、修改旧 Visio 图或多轮修订时使用。场景起图以 .scene.json 迭代，旧图以当前 .vsdx 迭代；不要求 Visio 的系统框图和科研配图交 research-schematic-imagegen，飞书文档或画板里的图交 lark-whiteboard，论文投稿图审查交 paper-figure-review。","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fvisiomaster\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fvisiomaster\u002FSKILL.md",[291],{"categoryNo":292,"categoryName":293,"taxonomyType":22,"taxonomyTypeDesc":23},"01JBSEEDCAT000000000000021","资料检索与研究",[286,295,296,297,298,299,300,301,302,303,304,305,306,307,308],"schematic","review","imagegen","image","生成","架构图","流程图","文字","shape","json","paper","research","windows","id",{"skillNo":310,"name":311,"nameZh":6,"description":312,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":313,"sourceHtmlUrl":314,"categories":315,"keywords":316,"publishedAt":52},"01M71F6VZN093VVJD7NCT809J9","skill-creator","创建、修改和改进技能，检查技能结构、触发说明及执行效果。根据实际影响选择文字检查、独立行为试跑或新旧对照，也提供按需使用的盲评、查看器、Claude 触发优化和打包工具。用户要求新建技能、调整流程、修复技能行为或比较技能效果时使用；普通业务任务不因此进入技能维护。","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fskill-creator\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fskill-creator\u002FSKILL.md",[],[317,318,277,302],"creator","skill creator",{"skillNo":320,"name":321,"nameZh":6,"description":322,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":241,"licenseFileUrl":323,"sourceHtmlUrl":324,"categories":325,"keywords":331,"publishedAt":52},"01M71F5XZPZ7Y7YYNFSAEFSQXF","paper-search","为科研问题检索、扩展并核实论文证据，也负责正式参考文献生成、DOI\u002F元数据一致性核验和已有引用纠错。Use when 用户要找论文、查文献、补参考文献、找代表作\u002F最新进展\u002F相反证据，或说“继续找”“再深入”；正式引用任务核实相关条目的出版字段，复用可定位且未变化的核验结果，有具体疑点只核相关条目。仅 BibTeX 语法、样式、引用键对应和编译问题转 latex-paper，不因 IEEE\u002FBibTeX 关键词全查。普通检索快速返回可核实且去重的结果，联网用 web-access。已知论文下载转 paper-download，原文技术核对和审稿转 paper-review，整篇稿件引用覆盖检查转对应论文编辑或审查技能。普通参数或多渠道比较表转 xlsx，明确持续产品库或库契约审计转 product-research-workbook，远程写回转对应文档技能。","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fpaper-search\u002FLICENSE.md","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fpaper-search\u002FSKILL.md",[326,327,328],{"categoryNo":25,"categoryName":26,"taxonomyType":22,"taxonomyTypeDesc":23},{"categoryNo":292,"categoryName":293,"taxonomyType":22,"taxonomyTypeDesc":23},{"categoryNo":329,"categoryName":330,"taxonomyType":30,"taxonomyTypeDesc":31},"01JBSEEDCAT000000000000031","产品经理",[305,332,333,334,335,299,33,336,44,337,263,338,339,274,306,340,296,341,265,273],"paper search","search","product-research","ieee","adaptation","evidence","product","数据","doi","关键词",{"skillNo":343,"name":344,"nameZh":6,"description":345,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":241,"licenseFileUrl":346,"sourceHtmlUrl":347,"categories":348,"keywords":351,"publishedAt":52},"01M71F5M874YA5YD9893SHFC6Y","paper-download","获取、复用或导入论文 PDF 到 Zotero，先查已有条目、附件和任务本地资料，只补请求的缺项；归档保留已有阅读内容与外部新增附件，超时先核实原操作。Use when 用户给出题名、DOI、论文 URL、作者主页、论文列表、本地 PDF 或 paper-search 结果，要求下载、保存到 Zotero、整理已有论文或继续机构授权下载。研究问题检索与引用元数据核验用 paper-search；原文技术核对用 paper-review。联网由 web-access 路由；Windows 版机构模块通过 web-access HTTP Broker 复用当前 Edge 会话，负责出版物核验，不自行执行浏览器登录或验证码。","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fpaper-download\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FSanAntonio021\u002Fagents-skills\u002Fblob\u002Fmain\u002Fpaper-download\u002FSKILL.md",[349,350],{"categoryNo":20,"categoryName":21,"taxonomyType":22,"taxonomyTypeDesc":23},{"categoryNo":292,"categoryName":293,"taxonomyType":22,"taxonomyTypeDesc":23},[352,353,305,354,340,48,307,276,339,355,296,282,356,265,273,357,274,358,333,263],"paper download","download","broker","整理","URL","http","Zotero"]