[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6JTGH5QK0AE3DVZRG519EES":3,"skill-same:01M6JTGH5QK0AE3DVZRG519EES":74},{"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":6,"sourceHtmlUrl":16,"categories":17,"keywords":23,"publishedAt":45,"linkOnly":46,"markdownContent":47,"markdownBody":48,"frontMatter":49,"refName":50,"commitSha":51,"skillDirectory":52,"forksCount":53,"repositoryHtmlUrl":54,"repositoryHomepageUrl":55,"ownerHtmlUrl":56,"ownerBlogUrl":57,"ownerDescription":58,"updatedAt":59,"install":60,"resources":73},"01M6JTGH5QK0AE3DVZRG519EES","using-bgs-translator",null,"Use when the user wants to translate a Bethesda Game Studios plugin (.esp\u002F.esm\u002F.esl) with bgs-translator\u002Fxtl, create or inspect a translation project, run LLM batch translation, generate prompt previews, or export SST\u002FXML for xTranslator. Triggers - \"translate this mod\", \"汉化这个 mod\", \"localize plugin\", \"build SST\", \"use xtl\", \"bgs-translator\".","hashgraph-online\u002Fawesome-codex-plugins","hashgraph-online","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F160371962?v=4","01M6CNB6NRJVNNGN14TR01Q6D3","HOL (Hashgraph Online)","01M6JT1HSKKG9V68KQAQGBA7XP",1575,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002FBB-84C\u002Fbgs-modding-superpowers\u002Fskills\u002Fusing-bgs-translator\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44],"using bgs translator","translator","export","when","for","inspect","esl","sst","triggers","project","localize","translation","translate","use","prompt","game","ESP","batch","run","generate","create","2026-11-10 10:12:30",false,"---\nname: using-bgs-translator\ndescription: Use when the user wants to translate a Bethesda Game Studios plugin (.esp\u002F.esm\u002F.esl) with bgs-translator\u002Fxtl, create or inspect a translation project, run LLM batch translation, generate prompt previews, or export SST\u002FXML for xTranslator. Triggers - \"translate this mod\", \"汉化这个 mod\", \"localize plugin\", \"build SST\", \"use xtl\", \"bgs-translator\".\n---\n\n# Using bgs-translator\n\n`bgs-translator` is the agent-driven translation pipeline for Bethesda plugin\ntext. It reads plugin strings, builds AI translation batches with glossary and\nprotected-span handling, writes a project memory database, and exports SST\u002FXML\ndictionaries for xTranslator or ESP-ESM Translator to finalize.\n\nThe browser GUI is the only GUI surface. The agent should prefer `xtl` for work\nit can do directly; use the GUI for human review, provider\u002Fkey setup when the\nuser wants it, prompt preview, progress monitoring, and manual cleanup.\n\n## Prerequisites — verify before any other command\n\n`xtl` is the CLI entry point. It is PyPI-distributed (`bgs-translator` package)\nand is **not** bundled with this plugin; the vendored tree only carries\ndocumentation and helper scripts. On a fresh machine `xtl version` will fail.\n\n**This is a long-tail install task that the agent owns**, not the user. Do not\nask the user to install Python or pipx by hand. Run the readback first; if it\nfails, route to the setup skill which performs the install and PATH-fix.\n\n```powershell\n# Readback. Run this BEFORE any of the workflow examples below.\nxtl version\n```\n\nIf `xtl version` errors with \"command not found\" or \"not recognized\", switch to\n`setting-up-bgs-modding-environment` (look at \"Step 3A - Install or verify\n`xtl`\" — it owns `pipx install bgs-translator==0.9.0rc1`, the pipx PATH-warning\nflow, and the post-install smoke). After install, return here and proceed.\n\nFor an already-installed `xtl` that is broken, stale, or behind a recent\nrelease, switch to `maintaining-modding-environments` (\"Translator CLI\nmaintenance\" section).\n\nThe KB cache that `xtl` reads from is `~\u002F.bgs-modding-superpowers\u002Fkb\u002Fpacks`,\nunified with this plugin's `bgs_kb_*` MCP tools. If `bgs_kb_status` lists a pack\nbut `xtl batch plan` still produces a single-entry glossary subset, suspect a\nstale legacy cache at `%LOCALAPPDATA%\u002Fbgs-modding-superpowers\u002Fkb` and surface it\nto the user instead of accepting the poor glossary silently.\n\nFor RAG-quality translation against a game, the user needs the per-game pack\nAND the corresponding localization pack (e.g. `bgs-kb-starfield` and\n`bgs-l10n-starfield-zhhans` for Starfield Simplified Chinese). Setup skill\nhandles both. If only one is installed, surface the gap before running real\nbatches.\n\nHuman manuals, read these files and explain them to users who need GUI help:\n\n- Chinese: `tools\u002Fbgs-translator\u002FUSER-GUIDE.zh-cn.md`\n- English: `tools\u002Fbgs-translator\u002FUSER-GUIDE.en.md`\n\nThese files ship inside the materialized plugin tree under\n`tools\u002Fbgs-translator\u002F` alongside `scripts\u002Frestart-web-gui.ps1`; they are\npresent in a fresh vendor-clone install.\n\n## 本地化判断 \u002F Localization judgment\n\nLocalization is not \"send every string to a provider.\" It is the curator's job\nto decide which player-visible text should change, which technical spans must\nstay exact, and which terms must remain consistent across the whole pack. A\ntranslation tool can produce candidate text; it cannot decide the pack glossary,\nthe winning string provider, or whether a string is actually safe to edit.\n\nThe core mechanic is the runtime string chain: the player sees the value from\nthe plugin\u002Fpatch layer that wins at runtime. A base-game translation or an older\nmod translation can be overridden by a later mod, update patch, or compatibility\npatch. Do not treat missing Chinese as a mysterious tool failure. First identify\nthe winning string provider, then localize that layer.\n\nSource basis: BB84's FO4 localization 疑难杂症 transcript (`[E8]` in\n`.opencode\u002Fartifacts\u002Fsixiang-build\u002Finjection-using-bgs-translator\u002Fframework-extraction.md`)\nexplains the override chain, attribution workflow, patch-localization ordering,\nand non-translatable\u002Fsignature caveats. That extraction also marks current\n`[GAP]` items; do not fill those gaps from guesswork.\n\n```dot\ndigraph localization_judgment {\n  rankdir=TB;\n  node [shape=box];\n\n  start [shape=doublecircle, label=\"String candidate\"];\n  visible [shape=diamond, label=\"Player-visible text?\\n(dialogue, quest, item name, book, UI, MCM, loading screen)\"];\n  protected [shape=diamond, label=\"Protected span or internal reference?\\nplaceholder, EditorID, script key, FormID, path, token\"];\n  kb [label=\"Query KB for this game's protected signatures\u002Ffields\\nand localization gotchas\"];\n  winning [label=\"Find the winning runtime provider\\nbase mod? update patch? compatibility patch? loose text file?\"];\n  glossary [shape=diamond, label=\"Term already locked by pack glossary?\"];\n  translate [label=\"Translate\u002Freuse comparison translation\\nwhile preserving protected spans\"];\n  normalize [label=\"Normalize to pack glossary\\nor record a deliberate term change\"];\n  review [label=\"Sample readback in context\\nneeds_review is real work\"];\n  lock [shape=doublecircle, label=\"LOCK \u002F do not translate\"];\n  skip [shape=doublecircle, label=\"SKIP \u002F not a localization target\"];\n  done [shape=doublecircle, label=\"ACCEPT localized string\"];\n\n  start -> visible;\n  visible -> skip [label=\"no\"];\n  visible -> protected [label=\"yes\"];\n  protected -> lock [label=\"yes\"];\n  protected -> kb [label=\"unclear \u002F no\"];\n  kb -> winning -> glossary;\n  glossary -> normalize [label=\"yes, differs\"];\n  glossary -> translate [label=\"no \u002F already matches\"];\n  normalize -> review;\n  translate -> review;\n  review -> done [label=\"context and placeholders pass\"];\n  review -> translate [label=\"mismatch\"];\n}\n```\n\n### Protected-span red flags\n\nNever localize a span just because it appears in an extracted text table. The\nsafe default is: player-facing prose can be translated; identifiers and engine\nreferences stay exact unless the per-game KB or the tool's schema proves they\nare display values.\n\n| Red flag | What to do instead |\n|---|---|\n| Placeholder, formatter token, alias marker, or bracket tag appears inside the text. | Preserve it byte-for-byte and translate only the surrounding prose. |\n| The value looks like an EditorID, script property, config key, JSON key, file path, plugin name, FormID, or hex ID. | Lock it unless KB\u002Ftool docs identify it as display text. |\n| A record\u002Fsignature\u002Ffield type is unfamiliar. | Query KB for the current game before translating; if KB is silent, mark `[GAP]` and review manually. |\n| Generated naming-rule or FormID-dependent text is selected for comparison\u002Fbulk translation. | Do not bulk-translate. Use manual semantic review or leave locked until KB confirms the safe handling. |\n| A translated plugin is being sorted after update\u002Fcompatibility patches to \"win\" text. | Do not break the functional patch layer. Localize the patch that actually wins, then load that localized patch after the functional layer. |\n\n### Rationalizations\n\n| Excuse | Reality |\n|---|---|\n| \"There must be a universal localization patch.\" | There is no final patch that covers every later mod-side string. Find the winning provider. |\n| \"The base-game Chinese patch is installed, so English text means the translator failed.\" | A later mod or patch may be overriding the localized base string. Attribute first. |\n| \"Just translate everything; the validator will catch problems.\" | Validators catch known placeholders, not every engine reference, naming-rule dependency, or glossary drift. |\n| \"The old Chinese ESP worked before; put it at the end.\" | That can invalidate update\u002Fcompatibility patches, and stale translations may miss new records. Regenerate against the current winning layer. |\n| \"Different mods can translate the same term differently; players will understand.\" | A pack needs one glossary. A later competing term is a judgment call, not accidental drift. |\n\n### KB query discipline\n\nThis section is game-agnostic. Do not fossilize FO4, Skyrim, or Starfield\nprotected-string lists in this skill. Before a real batch, query KB for the\ncurrent game and mod shape:\n\n```text\nbgs_kb_query({\n  query: \"localization protected strings placeholders EditorID FormID generated names\",\n  domains: [\"install-planning\", \"plugin-format\"],\n  games: [\"\u003Ccurrent game>\"]\n})\n\nbgs_kb_query({\n  query: \"translation glossary consistency string lookup override chain\",\n  domains: [\"install-planning\", \"load-order\"],\n  games: [\"\u003Ccurrent game>\"]\n})\n```\n\nIf KB is silent for a protected-string category, write `[GAP]` in the plan and\nkeep the string out of unattended AI translation. Per-game protected-string and\nFormID-dependency patterns belong in KB records, not in this skill body.\n\n## What xtl Does Better\n\nMost translator tools focus on editing string tables. `xtl` builds a complete\nLLM request for each batch:\n\n1. It extracts only translatable plugin text into project memory.\n2. It masks dangerous placeholders such as `{P0}` and `\u003CAlias=...>`.\n3. It deduplicates repeated text where safe.\n4. It recalls relevant game\u002Fmod\u002Fplayer terminology.\n5. It adds game context, mod context, signature explanations, style rules, and\n   do-not-translate rules to a system prompt.\n6. It sends compact JSON-shaped work items to the selected provider.\n7. It validates the response before writing translations back to project memory.\n8. It exports SST\u002FXML for xTranslator instead of editing the original plugin.\n\nThis is why the output can be more context-aware than plain string-table machine\ntranslation: the LLM receives the text plus the mod's purpose, game lore,\nrecord-type meaning, glossary, and placeholder rules in one request.\n\n## Help and Capability Discovery\n\nNever assume a shown command is the only valid shape. Examples below are\ntemplates. Before using an unfamiliar option, check the live CLI:\n\n```powershell\n# Show all top-level xtl command groups.\nxtl --help\n\n# Show provider commands and accepted provider options.\nxtl profile --help\nxtl profile add --help\nxtl profile edit --help\n\n# Show project import\u002Fexport options.\nxtl project --help\nxtl project init --help\nxtl project export --help\n\n# Show inspect, planning, run, status, and cancellation options.\nxtl inspect --help\nxtl batch plan --help\nxtl batch run --help\nxtl batch status --help\n```\n\nProvider `--sdk-kind` is not always `openai-compat`. Check `xtl profile add\n--help` for the currently supported values. At the time of this skill, the\ncode supports `openai`, `anthropic`, `gemini`, and `openai-compat`.\n\n## Operating Rules\n\n- Do not edit `.esp`, `.esm`, or `.esl` directly. `xtl` emits SST\u002FXML only.\n- Do not route translator work through xEdit or MO2 unless the user explicitly\n  asks for an unrelated xEdit\u002FMO2 task.\n- Do not accept API keys in chat or command history. Use `xtl profile set-key`\n  for hidden interactive input, or let the user enter the key in the browser GUI.\n- Do not bypass prompt preview when the user has enabled it. If the user asks for\n  a fully agent-run batch, set\u002Fconfirm `behavior.prompt_preview_required=false`\n  before dispatching.\n- Always surface plan size, batch count, skipped placeholder-only count, and the\n  output dictionary path.\n- Treat `needs_review` as real work, not as success. Either fix it or report it.\n\n## CLI End-to-End Flow\n\nUse this flow when the user wants the agent to run the translation process.\n\n1. Verify the local CLI.\n\n   ```powershell\n   # Confirm the installed\u002Fcurrent xtl can start and report capabilities.\n   xtl version\n   ```\n\n2. Inspect the plugin before creating a project.\n\n   ```powershell\n   # Example path only. Replace with the user's real ESP\u002FESM\u002FESL.\n   xtl inspect plugin \"D:\\path\\to\\Mod.esm\"\n   ```\n\n   If game detection is ambiguous, pass a real game value. Do not guess silently.\n   Check available options with `xtl inspect plugin --help`.\n\n3. Create or refresh a project.\n\n   ```powershell\n   # Create project memory from the source plugin. Example target language: zh-cn.\n   xtl project init \u003Cproject> --plugin \"D:\\path\\to\\Mod.esm\" --target-lang zh-cn\n   ```\n\n4. Configure the provider if needed.\n\n   ```powershell\n   # Example: OpenAI-compatible provider. Check xtl profile add --help for all sdk kinds.\n   xtl profile add \u003Cprofile> --sdk-kind openai-compat --base-url \u003Capi-root> --model \u003Cmodel> --api-key-env \u003CENV_NAME> --json-mode json_object\n\n   # Store the key with hidden input. Never put the real key in the command line.\n   xtl profile set-key \u003Cprofile>\n\n   # Example advanced settings. Use xtl profile edit --help before changing these.\n   xtl profile edit \u003Cprofile> --max-concurrency 8 --rate-limit-rpm 120 --rate-limit-tpm 90000\n\n   # Test the provider and then make it active.\n   xtl profile probe \u003Cprofile>\n   xtl profile activate \u003Cprofile>\n   ```\n\n   `set-key` prompts with hidden input and writes `translator\u002Fprofiles\u002F.env`.\n   Never place a real key in a shell command, commit, log, or chat message.\n\n5. Inspect project content.\n\n   ```powershell\n   # Get high-level counts by record signature.\n   xtl inspect signatures \u003Cproject>\n\n   # Example: inspect 100 quest entries. Check xtl inspect entries --help for filters.\n   xtl inspect entries \u003Cproject> --sig QUST --limit 100\n   ```\n\n6. Research context before planning. For real mods, gather:\n\n   - game-world context relevant to the mod;\n   - mod context from the mod page\u002Freadme;\n   - sampled entries by signature\u002Ffield;\n   - protected placeholders and recurring tags that must remain unchanged.\n\n7. Plan the batch. For \"all currently untranslated\" use filters; for GUI\n   queue\u002Ffanatic mode use `--queue \u003Cqueue_id>` when the user submitted one.\n\n   ```powershell\n   # Example: plan untranslated QUST text in groups of 200.\n   # This only writes plan.json and does not call the provider.\n   xtl batch plan \u003Cproject> --register dialogue --target-lang zh-cn --profile \u003Cprofile> --batch-size 200 --sig QUST --game-lore-world \"Starfield 2330 Settled Systems\" --game-lore-summary \"\u003Cdetailed lore>\" --mod-name \"\u003Cmod name>\" --mod-theme \"\u003Cdetailed mod context>\" --style \"Polished Simplified Chinese game localization.\"\n   ```\n\n   `--register`, `--sig`, `--field`, batch size, and context fields are examples.\n   Check `xtl batch plan --help` and inspect project signatures before choosing.\n   Report `plan_id`, `total_items`, `batch_count`, `skipped_reasons`, and the\n   plan path.\n\n8. Dispatch the run.\n\n   ```powershell\n   # Start a background worker and return immediately with run_id and log paths.\n   xtl batch run \u003Cproject> --plan \u003Cplan_id>\n   ```\n\n   `xtl batch run` returns immediately by default. Use `--wait` only for tests or\n   deliberate foreground execution. Use `--dry-run` only for smoke tests. For\n   no-human-preview agent runs, first confirm:\n\n   ```powershell\n   # Disable required browser prompt preview only when the user asked for agent-only execution.\n   xtl config set behavior.prompt_preview_required false\n   ```\n\n9. Poll status and logs.\n\n   ```powershell\n   # Poll periodically after background launch.\n   Start-Sleep -Seconds 10\n   xtl batch status \u003Crun_id>\n\n   # Inspect recent persisted run files\u002Flogs if progress looks stuck.\n   xtl batch logs \u003Crun_id>\n\n   # Request cancellation if the user asks to stop.\n   xtl batch cancel \u003Crun_id>\n   ```\n\n10. Validate and export.\n\n    ```powershell\n    # Validate project state before export.\n    xtl validate project \u003Cproject>\n\n    # Export xTranslator dictionary output.\n    xtl project export \u003Cproject> --format sst\n    ```\n\n    Tell the user which SST\u002FXML files were emitted and that xTranslator\u002FESP-ESM\n    Translator owns the final \"Finalize\" step.\n\n## xTranslator Batch Processor Handoff\n\nUse this when the user wants the agent to prepare xTranslator finalization but\naccepts that a human must click through xTranslator. The human's job is only:\nlaunch xTranslator through MO2 when needed, open `Wizards -> Batch Processor`,\npaste the script provided by the agent, run it, then copy the bottom `Log` tab\nback to the agent. The agent must inspect the log for every `StartRule`:\n\n- `Loading Esp\u002Fesm\u002Fesl file...` succeeds for the intended file.\n- `Applying sstRessources...` and `importing\u002Fapplying vocabfile` run.\n- `Saving Esp\u002FEsm\u002FEsl files...` or `Saving Strings files...` appears.\n- No `BatchProcessor: Error` line appears.\n- Saved paths point at the intended `- SC` output, not the source mod unless a\n  deliberate loose-strings staging fallback was used.\n\nxTranslator's command line cannot run this workflow headlessly. It can load one\nfile, but SST import\u002Ffinalize is driven by Batch Processor text.\n\nBefore generating any Batch Processor script, create or verify the corresponding\nMO2 output mod folder under `...\\MO2\\mods\\` using the source mod's name plus\n` - SC`, for example `D:\\Starfield MO2\\mods\\Some Mod - SC`. This folder is the\nonly intended landing place for finalized translation files. Do not let\nxTranslator save final deliverables into the source mod unless using the\ntemporary loose-strings staging fallback described below.\n\n```powershell\n# Create the sibling MO2 mod that will receive finalized Chinese output.\nNew-Item -ItemType Directory -Force -Path \"D:\\Starfield MO2\\mods\\Some Mod - SC\" | Out-Null\n```\n\n### Batch Processor language\n\nGenerate one `StartRule` block per plugin. Use full Windows paths. Keep path\narguments unquoted inside the Batch Processor block; xTranslator reads the rest\nof the line after the command prefix, so spaces are valid there.\n\nImportant commands:\n\n- `LangSource=en` and `LangDest=zhhans` select the language pair.\n- `UseDataDir=false` tells xTranslator to use the explicit file path instead of\n  the game Data folder.\n- `Command=LoadFile:\u003Cplugin>` loads one `.esm\u002F.esp\u002F.esl`.\n- `Command=ImportSst:\u003CapplyOn>:\u003Cmode>:\u003Csst>` imports an SST dictionary.\n- `Command=Finalize` writes the final output for the loaded file.\n- `Command=CloseFile` closes it before the next rule.\n\nUseful `ImportSst` parameters:\n\n- `applyOn=0`: apply to everything.\n- `applyOn=1`: apply only to not-translated items.\n- `mode=1`: strict FormID + original string matching.\n- `mode=3`: strings-only matching.\n\nPreferred pattern: run strict first, then optionally use strings-only only for\nstill-untranslated items:\n\n```text\nCommand=ImportSst:0:1:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=ImportSst:1:3:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\n```\n\nIf the user is debugging record identity, omit the strings-only fallback so\nunmatched records remain visible.\n\n### Case A - plugin-finalized output\n\nUse this for mods where xTranslator saves the plugin itself, such as a compact\nESM\u002FESP\u002FESL workflow. Do not load the source plugin. First copy the source\nplugin to the sibling MO2 overlay directory named `Original Mod Name - SC`, then\nload that copied file in Batch Processor.\n\n```powershell\n# Prepare the sibling MO2 overlay so xTranslator finalizes a copy, not the source plugin.\nNew-Item -ItemType Directory -Force -Path \"D:\\path\\to\\Mod Name - SC\" | Out-Null\nCopy-Item -LiteralPath \"D:\\path\\to\\Mod Name\\mod.esm\" -Destination \"D:\\path\\to\\Mod Name - SC\\mod.esm\" -Force\n```\n\nThen provide the human this Batch Processor block:\n\n```text\nStartRule\nLangSource=en\nLangDest=zhhans\nUseDataDir=false\nCommand=LoadFile:D:\\path\\to\\Mod Name - SC\\mod.esm\nCommand=ImportSst:0:1:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=ImportSst:1:3:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=Finalize\nCommand=CloseFile\nEndRule\n```\n\nExpected success log contains `Saving Esp\u002FEsm\u002FEsl files...` and a saved path\nunder `Mod Name - SC`.\n\n### Case B - loose Strings output\n\nUse this for string-backed Starfield plugins where xTranslator finalizes\n`strings\\*_zhhans.strings`, `*.dlstrings`, and `*.ilstrings` instead of changing\nthe BA2. Preferred: copy the plugin and text BA2 to `- SC`, load the copied\nplugin there, and let xTranslator write loose strings directly under `- SC`.\n\n```powershell\n# Prepare a sibling overlay with the plugin and the BA2 that contains source strings.\nNew-Item -ItemType Directory -Force -Path \"D:\\path\\to\\Mod Name - SC\" | Out-Null\nCopy-Item -LiteralPath \"D:\\path\\to\\Mod Name\\mod.esm\" -Destination \"D:\\path\\to\\Mod Name - SC\\mod.esm\" -Force\nCopy-Item -LiteralPath \"D:\\path\\to\\Mod Name\\mod - main.ba2\" -Destination \"D:\\path\\to\\Mod Name - SC\\mod - main.ba2\" -Force\n```\n\nBatch Processor block:\n\n```text\nStartRule\nLangSource=en\nLangDest=zhhans\nUseDataDir=false\nCommand=LoadFile:D:\\path\\to\\Mod Name - SC\\mod.esm\nCommand=ImportSst:0:1:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=ImportSst:1:3:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=Finalize\nCommand=CloseFile\nEndRule\n```\n\nExpected success log contains `Saving Strings files...` and saved files under\n`Mod Name - SC\\strings\\`.\n\nAfter verifying the target-language loose files exist under `Mod Name -\nSC\\strings\\`, delete the temporary copied plugin\u002Farchive files from `Mod Name -\nSC`. For string-localized mods, the delivered MO2 overlay only needs the loose\n`strings\\*_zhhans.strings`, `*.dlstrings`, and `*.ilstrings` files; keeping the\ncopied `.esm\u002F.esp\u002F.esl`, `.ba2`, or `.bsa` in `- SC` can shadow the source mod\nunnecessarily.\n\n```powershell\n# After successful string finalization, keep only the loose translated strings.\nRemove-Item -LiteralPath \"D:\\path\\to\\Mod Name - SC\\mod.esm\" -Force\nRemove-Item -LiteralPath \"D:\\path\\to\\Mod Name - SC\\mod - main.ba2\" -Force\n```\n\nIf the BA2 is too large to copy and the user accepts a staging step, load the\nsource plugin and let xTranslator write loose target-language strings into the\nsource mod's `strings` folder. Immediately move only the generated target\nlanguage files to `- SC\\strings`:\n\n```powershell\n# Move the loose target-language strings produced by xTranslator into the sibling overlay.\nNew-Item -ItemType Directory -Force -Path \"D:\\path\\to\\Mod Name - SC\\strings\" | Out-Null\nMove-Item -LiteralPath \"D:\\path\\to\\Mod Name\\strings\\mod_zhhans.strings\" -Destination \"D:\\path\\to\\Mod Name - SC\\strings\\mod_zhhans.strings\" -Force\nMove-Item -LiteralPath \"D:\\path\\to\\Mod Name\\strings\\mod_zhhans.dlstrings\" -Destination \"D:\\path\\to\\Mod Name - SC\\strings\\mod_zhhans.dlstrings\" -Force\nMove-Item -LiteralPath \"D:\\path\\to\\Mod Name\\strings\\mod_zhhans.ilstrings\" -Destination \"D:\\path\\to\\Mod Name - SC\\strings\\mod_zhhans.ilstrings\" -Force\n```\n\nOnly use this fallback when the user explicitly accepts temporary writes to the\nsource mod directory. Never write into the game Stock\u002FData directory.\n\n## Browser GUI Flow\n\nUse the GUI when a human needs to configure, inspect, approve, or manually fix:\n\n```powershell\n# Launch the local browser control panel.\nxtl gui\n```\n\nIf the GUI or its process state becomes inconsistent, restart it with the stable\nhelper:\n\n```powershell\n# From tools\u002Fbgs-translator, restart the web GUI on the usual local port.\npowershell -ExecutionPolicy Bypass -File .\\scripts\\restart-web-gui.ps1 -Port 7847\n```\n\nThe GUI is browser-only. There is no Tk fallback.\n\nHigh-signal tabs:\n\n- Project: import plugin projects, see project metadata, export SST, open export\n  directory, and see record signature translation stats.\n- Entries: filter, inspect source\u002Fdestination, quick translate, multi-select,\n  submit selected entries, or use fanatic mode for all filtered entries.\n- Prompt: inspect current pending prompt and approve\u002Fapprove-all\u002Fskip\u002Fdiscard.\n- Batches: watch run progress, request stop, inspect historical audit records,\n  and discard completed run translations when supported.\n- Profiles: create\u002Factivate\u002Fprobe providers and store API keys locally.\n- Glossary: edit player terms and do-not-translate terms.\n- Logs: inspect run files and technical logs.\n\nFor detailed player-facing guidance, point the user to the manuals listed at the\ntop of this skill.\n\nFor maintaining Starfield official terminology packs, mod-specific glossary\npacks, or third-party Skyrim\u002FFallout localization KBs, switch to\n`maintaining-modding-environments`. Translator runs consume those packs; they do\nnot own KB release or provenance policy.\n\n## Prompt Context Requirements\n\nDo not send generic prompts. For every real batch, build context with:\n\n- the game, era, faction\u002Fsetting, and localization style;\n- the mod's purpose, systems, characters, places, terminology, and UI tone;\n- signature explanations only for signatures present in the batch;\n- player glossary and do-not-translate rules;\n- protected placeholder rules such as `{P0}` and `\u003CAlias=...>`.\n\nIf the user provides a mod URL, browse it and summarize relevant features. If\nthe mod page is unavailable, use local readmes\u002Fproject entries and state the\ngap.\n\n## Status Semantics\n\nAlign user-facing status with xTranslator concepts:\n\n- `untranslated`: no accepted translation exists.\n- `translated`: validated\u002Faccepted translation.\n- `needs_review`: partial or suspicious translation that requires human review.\n- `locked`: should remain unchanged and should not be sent to AI.\n\nPlaceholder-only and number-only strings should not enter AI batches. Do not\nhide a mismatch between GUI counts and exported SST; investigate before\nclaiming success.\n\n## Anti-Patterns\n\n- Do not copy examples blindly. Check `--help` for supported values first.\n- Do not make up hard caps. Respect user batch size\u002Fbudget settings; if a prompt\n  may exceed context, report it and ask for smaller batches\u002Fsettings.\n- Do not mark English source copied to destination as translated unless it is a\n  deliberate locked\u002Fno-translate entry.\n- Do not rely on string-only SST matching as proof of correct export. FormID and\n  record identity should match xTranslator where possible.\n- Do not leave orphaned pending-preview state after a service restart. Historical\n  plans are audit records unless a real resume command exists.\n- Do not promise cancellation is free; already-started provider calls may bill.\n","# Using bgs-translator\n\n`bgs-translator` is the agent-driven translation pipeline for Bethesda plugin\ntext. It reads plugin strings, builds AI translation batches with glossary and\nprotected-span handling, writes a project memory database, and exports SST\u002FXML\ndictionaries for xTranslator or ESP-ESM Translator to finalize.\n\nThe browser GUI is the only GUI surface. The agent should prefer `xtl` for work\nit can do directly; use the GUI for human review, provider\u002Fkey setup when the\nuser wants it, prompt preview, progress monitoring, and manual cleanup.\n\n## Prerequisites — verify before any other command\n\n`xtl` is the CLI entry point. It is PyPI-distributed (`bgs-translator` package)\nand is **not** bundled with this plugin; the vendored tree only carries\ndocumentation and helper scripts. On a fresh machine `xtl version` will fail.\n\n**This is a long-tail install task that the agent owns**, not the user. Do not\nask the user to install Python or pipx by hand. Run the readback first; if it\nfails, route to the setup skill which performs the install and PATH-fix.\n\n```powershell\n# Readback. Run this BEFORE any of the workflow examples below.\nxtl version\n```\n\nIf `xtl version` errors with \"command not found\" or \"not recognized\", switch to\n`setting-up-bgs-modding-environment` (look at \"Step 3A - Install or verify\n`xtl`\" — it owns `pipx install bgs-translator==0.9.0rc1`, the pipx PATH-warning\nflow, and the post-install smoke). After install, return here and proceed.\n\nFor an already-installed `xtl` that is broken, stale, or behind a recent\nrelease, switch to `maintaining-modding-environments` (\"Translator CLI\nmaintenance\" section).\n\nThe KB cache that `xtl` reads from is `~\u002F.bgs-modding-superpowers\u002Fkb\u002Fpacks`,\nunified with this plugin's `bgs_kb_*` MCP tools. If `bgs_kb_status` lists a pack\nbut `xtl batch plan` still produces a single-entry glossary subset, suspect a\nstale legacy cache at `%LOCALAPPDATA%\u002Fbgs-modding-superpowers\u002Fkb` and surface it\nto the user instead of accepting the poor glossary silently.\n\nFor RAG-quality translation against a game, the user needs the per-game pack\nAND the corresponding localization pack (e.g. `bgs-kb-starfield` and\n`bgs-l10n-starfield-zhhans` for Starfield Simplified Chinese). Setup skill\nhandles both. If only one is installed, surface the gap before running real\nbatches.\n\nHuman manuals, read these files and explain them to users who need GUI help:\n\n- Chinese: `tools\u002Fbgs-translator\u002FUSER-GUIDE.zh-cn.md`\n- English: `tools\u002Fbgs-translator\u002FUSER-GUIDE.en.md`\n\nThese files ship inside the materialized plugin tree under\n`tools\u002Fbgs-translator\u002F` alongside `scripts\u002Frestart-web-gui.ps1`; they are\npresent in a fresh vendor-clone install.\n\n## 本地化判断 \u002F Localization judgment\n\nLocalization is not \"send every string to a provider.\" It is the curator's job\nto decide which player-visible text should change, which technical spans must\nstay exact, and which terms must remain consistent across the whole pack. A\ntranslation tool can produce candidate text; it cannot decide the pack glossary,\nthe winning string provider, or whether a string is actually safe to edit.\n\nThe core mechanic is the runtime string chain: the player sees the value from\nthe plugin\u002Fpatch layer that wins at runtime. A base-game translation or an older\nmod translation can be overridden by a later mod, update patch, or compatibility\npatch. Do not treat missing Chinese as a mysterious tool failure. First identify\nthe winning string provider, then localize that layer.\n\nSource basis: BB84's FO4 localization 疑难杂症 transcript (`[E8]` in\n`.opencode\u002Fartifacts\u002Fsixiang-build\u002Finjection-using-bgs-translator\u002Fframework-extraction.md`)\nexplains the override chain, attribution workflow, patch-localization ordering,\nand non-translatable\u002Fsignature caveats. That extraction also marks current\n`[GAP]` items; do not fill those gaps from guesswork.\n\n```dot\ndigraph localization_judgment {\n  rankdir=TB;\n  node [shape=box];\n\n  start [shape=doublecircle, label=\"String candidate\"];\n  visible [shape=diamond, label=\"Player-visible text?\\n(dialogue, quest, item name, book, UI, MCM, loading screen)\"];\n  protected [shape=diamond, label=\"Protected span or internal reference?\\nplaceholder, EditorID, script key, FormID, path, token\"];\n  kb [label=\"Query KB for this game's protected signatures\u002Ffields\\nand localization gotchas\"];\n  winning [label=\"Find the winning runtime provider\\nbase mod? update patch? compatibility patch? loose text file?\"];\n  glossary [shape=diamond, label=\"Term already locked by pack glossary?\"];\n  translate [label=\"Translate\u002Freuse comparison translation\\nwhile preserving protected spans\"];\n  normalize [label=\"Normalize to pack glossary\\nor record a deliberate term change\"];\n  review [label=\"Sample readback in context\\nneeds_review is real work\"];\n  lock [shape=doublecircle, label=\"LOCK \u002F do not translate\"];\n  skip [shape=doublecircle, label=\"SKIP \u002F not a localization target\"];\n  done [shape=doublecircle, label=\"ACCEPT localized string\"];\n\n  start -> visible;\n  visible -> skip [label=\"no\"];\n  visible -> protected [label=\"yes\"];\n  protected -> lock [label=\"yes\"];\n  protected -> kb [label=\"unclear \u002F no\"];\n  kb -> winning -> glossary;\n  glossary -> normalize [label=\"yes, differs\"];\n  glossary -> translate [label=\"no \u002F already matches\"];\n  normalize -> review;\n  translate -> review;\n  review -> done [label=\"context and placeholders pass\"];\n  review -> translate [label=\"mismatch\"];\n}\n```\n\n### Protected-span red flags\n\nNever localize a span just because it appears in an extracted text table. The\nsafe default is: player-facing prose can be translated; identifiers and engine\nreferences stay exact unless the per-game KB or the tool's schema proves they\nare display values.\n\n| Red flag | What to do instead |\n|---|---|\n| Placeholder, formatter token, alias marker, or bracket tag appears inside the text. | Preserve it byte-for-byte and translate only the surrounding prose. |\n| The value looks like an EditorID, script property, config key, JSON key, file path, plugin name, FormID, or hex ID. | Lock it unless KB\u002Ftool docs identify it as display text. |\n| A record\u002Fsignature\u002Ffield type is unfamiliar. | Query KB for the current game before translating; if KB is silent, mark `[GAP]` and review manually. |\n| Generated naming-rule or FormID-dependent text is selected for comparison\u002Fbulk translation. | Do not bulk-translate. Use manual semantic review or leave locked until KB confirms the safe handling. |\n| A translated plugin is being sorted after update\u002Fcompatibility patches to \"win\" text. | Do not break the functional patch layer. Localize the patch that actually wins, then load that localized patch after the functional layer. |\n\n### Rationalizations\n\n| Excuse | Reality |\n|---|---|\n| \"There must be a universal localization patch.\" | There is no final patch that covers every later mod-side string. Find the winning provider. |\n| \"The base-game Chinese patch is installed, so English text means the translator failed.\" | A later mod or patch may be overriding the localized base string. Attribute first. |\n| \"Just translate everything; the validator will catch problems.\" | Validators catch known placeholders, not every engine reference, naming-rule dependency, or glossary drift. |\n| \"The old Chinese ESP worked before; put it at the end.\" | That can invalidate update\u002Fcompatibility patches, and stale translations may miss new records. Regenerate against the current winning layer. |\n| \"Different mods can translate the same term differently; players will understand.\" | A pack needs one glossary. A later competing term is a judgment call, not accidental drift. |\n\n### KB query discipline\n\nThis section is game-agnostic. Do not fossilize FO4, Skyrim, or Starfield\nprotected-string lists in this skill. Before a real batch, query KB for the\ncurrent game and mod shape:\n\n```text\nbgs_kb_query({\n  query: \"localization protected strings placeholders EditorID FormID generated names\",\n  domains: [\"install-planning\", \"plugin-format\"],\n  games: [\"\u003Ccurrent game>\"]\n})\n\nbgs_kb_query({\n  query: \"translation glossary consistency string lookup override chain\",\n  domains: [\"install-planning\", \"load-order\"],\n  games: [\"\u003Ccurrent game>\"]\n})\n```\n\nIf KB is silent for a protected-string category, write `[GAP]` in the plan and\nkeep the string out of unattended AI translation. Per-game protected-string and\nFormID-dependency patterns belong in KB records, not in this skill body.\n\n## What xtl Does Better\n\nMost translator tools focus on editing string tables. `xtl` builds a complete\nLLM request for each batch:\n\n1. It extracts only translatable plugin text into project memory.\n2. It masks dangerous placeholders such as `{P0}` and `\u003CAlias=...>`.\n3. It deduplicates repeated text where safe.\n4. It recalls relevant game\u002Fmod\u002Fplayer terminology.\n5. It adds game context, mod context, signature explanations, style rules, and\n   do-not-translate rules to a system prompt.\n6. It sends compact JSON-shaped work items to the selected provider.\n7. It validates the response before writing translations back to project memory.\n8. It exports SST\u002FXML for xTranslator instead of editing the original plugin.\n\nThis is why the output can be more context-aware than plain string-table machine\ntranslation: the LLM receives the text plus the mod's purpose, game lore,\nrecord-type meaning, glossary, and placeholder rules in one request.\n\n## Help and Capability Discovery\n\nNever assume a shown command is the only valid shape. Examples below are\ntemplates. Before using an unfamiliar option, check the live CLI:\n\n```powershell\n# Show all top-level xtl command groups.\nxtl --help\n\n# Show provider commands and accepted provider options.\nxtl profile --help\nxtl profile add --help\nxtl profile edit --help\n\n# Show project import\u002Fexport options.\nxtl project --help\nxtl project init --help\nxtl project export --help\n\n# Show inspect, planning, run, status, and cancellation options.\nxtl inspect --help\nxtl batch plan --help\nxtl batch run --help\nxtl batch status --help\n```\n\nProvider `--sdk-kind` is not always `openai-compat`. Check `xtl profile add\n--help` for the currently supported values. At the time of this skill, the\ncode supports `openai`, `anthropic`, `gemini`, and `openai-compat`.\n\n## Operating Rules\n\n- Do not edit `.esp`, `.esm`, or `.esl` directly. `xtl` emits SST\u002FXML only.\n- Do not route translator work through xEdit or MO2 unless the user explicitly\n  asks for an unrelated xEdit\u002FMO2 task.\n- Do not accept API keys in chat or command history. Use `xtl profile set-key`\n  for hidden interactive input, or let the user enter the key in the browser GUI.\n- Do not bypass prompt preview when the user has enabled it. If the user asks for\n  a fully agent-run batch, set\u002Fconfirm `behavior.prompt_preview_required=false`\n  before dispatching.\n- Always surface plan size, batch count, skipped placeholder-only count, and the\n  output dictionary path.\n- Treat `needs_review` as real work, not as success. Either fix it or report it.\n\n## CLI End-to-End Flow\n\nUse this flow when the user wants the agent to run the translation process.\n\n1. Verify the local CLI.\n\n   ```powershell\n   # Confirm the installed\u002Fcurrent xtl can start and report capabilities.\n   xtl version\n   ```\n\n2. Inspect the plugin before creating a project.\n\n   ```powershell\n   # Example path only. Replace with the user's real ESP\u002FESM\u002FESL.\n   xtl inspect plugin \"D:\\path\\to\\Mod.esm\"\n   ```\n\n   If game detection is ambiguous, pass a real game value. Do not guess silently.\n   Check available options with `xtl inspect plugin --help`.\n\n3. Create or refresh a project.\n\n   ```powershell\n   # Create project memory from the source plugin. Example target language: zh-cn.\n   xtl project init \u003Cproject> --plugin \"D:\\path\\to\\Mod.esm\" --target-lang zh-cn\n   ```\n\n4. Configure the provider if needed.\n\n   ```powershell\n   # Example: OpenAI-compatible provider. Check xtl profile add --help for all sdk kinds.\n   xtl profile add \u003Cprofile> --sdk-kind openai-compat --base-url \u003Capi-root> --model \u003Cmodel> --api-key-env \u003CENV_NAME> --json-mode json_object\n\n   # Store the key with hidden input. Never put the real key in the command line.\n   xtl profile set-key \u003Cprofile>\n\n   # Example advanced settings. Use xtl profile edit --help before changing these.\n   xtl profile edit \u003Cprofile> --max-concurrency 8 --rate-limit-rpm 120 --rate-limit-tpm 90000\n\n   # Test the provider and then make it active.\n   xtl profile probe \u003Cprofile>\n   xtl profile activate \u003Cprofile>\n   ```\n\n   `set-key` prompts with hidden input and writes `translator\u002Fprofiles\u002F.env`.\n   Never place a real key in a shell command, commit, log, or chat message.\n\n5. Inspect project content.\n\n   ```powershell\n   # Get high-level counts by record signature.\n   xtl inspect signatures \u003Cproject>\n\n   # Example: inspect 100 quest entries. Check xtl inspect entries --help for filters.\n   xtl inspect entries \u003Cproject> --sig QUST --limit 100\n   ```\n\n6. Research context before planning. For real mods, gather:\n\n   - game-world context relevant to the mod;\n   - mod context from the mod page\u002Freadme;\n   - sampled entries by signature\u002Ffield;\n   - protected placeholders and recurring tags that must remain unchanged.\n\n7. Plan the batch. For \"all currently untranslated\" use filters; for GUI\n   queue\u002Ffanatic mode use `--queue \u003Cqueue_id>` when the user submitted one.\n\n   ```powershell\n   # Example: plan untranslated QUST text in groups of 200.\n   # This only writes plan.json and does not call the provider.\n   xtl batch plan \u003Cproject> --register dialogue --target-lang zh-cn --profile \u003Cprofile> --batch-size 200 --sig QUST --game-lore-world \"Starfield 2330 Settled Systems\" --game-lore-summary \"\u003Cdetailed lore>\" --mod-name \"\u003Cmod name>\" --mod-theme \"\u003Cdetailed mod context>\" --style \"Polished Simplified Chinese game localization.\"\n   ```\n\n   `--register`, `--sig`, `--field`, batch size, and context fields are examples.\n   Check `xtl batch plan --help` and inspect project signatures before choosing.\n   Report `plan_id`, `total_items`, `batch_count`, `skipped_reasons`, and the\n   plan path.\n\n8. Dispatch the run.\n\n   ```powershell\n   # Start a background worker and return immediately with run_id and log paths.\n   xtl batch run \u003Cproject> --plan \u003Cplan_id>\n   ```\n\n   `xtl batch run` returns immediately by default. Use `--wait` only for tests or\n   deliberate foreground execution. Use `--dry-run` only for smoke tests. For\n   no-human-preview agent runs, first confirm:\n\n   ```powershell\n   # Disable required browser prompt preview only when the user asked for agent-only execution.\n   xtl config set behavior.prompt_preview_required false\n   ```\n\n9. Poll status and logs.\n\n   ```powershell\n   # Poll periodically after background launch.\n   Start-Sleep -Seconds 10\n   xtl batch status \u003Crun_id>\n\n   # Inspect recent persisted run files\u002Flogs if progress looks stuck.\n   xtl batch logs \u003Crun_id>\n\n   # Request cancellation if the user asks to stop.\n   xtl batch cancel \u003Crun_id>\n   ```\n\n10. Validate and export.\n\n    ```powershell\n    # Validate project state before export.\n    xtl validate project \u003Cproject>\n\n    # Export xTranslator dictionary output.\n    xtl project export \u003Cproject> --format sst\n    ```\n\n    Tell the user which SST\u002FXML files were emitted and that xTranslator\u002FESP-ESM\n    Translator owns the final \"Finalize\" step.\n\n## xTranslator Batch Processor Handoff\n\nUse this when the user wants the agent to prepare xTranslator finalization but\naccepts that a human must click through xTranslator. The human's job is only:\nlaunch xTranslator through MO2 when needed, open `Wizards -> Batch Processor`,\npaste the script provided by the agent, run it, then copy the bottom `Log` tab\nback to the agent. The agent must inspect the log for every `StartRule`:\n\n- `Loading Esp\u002Fesm\u002Fesl file...` succeeds for the intended file.\n- `Applying sstRessources...` and `importing\u002Fapplying vocabfile` run.\n- `Saving Esp\u002FEsm\u002FEsl files...` or `Saving Strings files...` appears.\n- No `BatchProcessor: Error` line appears.\n- Saved paths point at the intended `- SC` output, not the source mod unless a\n  deliberate loose-strings staging fallback was used.\n\nxTranslator's command line cannot run this workflow headlessly. It can load one\nfile, but SST import\u002Ffinalize is driven by Batch Processor text.\n\nBefore generating any Batch Processor script, create or verify the corresponding\nMO2 output mod folder under `...\\MO2\\mods\\` using the source mod's name plus\n` - SC`, for example `D:\\Starfield MO2\\mods\\Some Mod - SC`. This folder is the\nonly intended landing place for finalized translation files. Do not let\nxTranslator save final deliverables into the source mod unless using the\ntemporary loose-strings staging fallback described below.\n\n```powershell\n# Create the sibling MO2 mod that will receive finalized Chinese output.\nNew-Item -ItemType Directory -Force -Path \"D:\\Starfield MO2\\mods\\Some Mod - SC\" | Out-Null\n```\n\n### Batch Processor language\n\nGenerate one `StartRule` block per plugin. Use full Windows paths. Keep path\narguments unquoted inside the Batch Processor block; xTranslator reads the rest\nof the line after the command prefix, so spaces are valid there.\n\nImportant commands:\n\n- `LangSource=en` and `LangDest=zhhans` select the language pair.\n- `UseDataDir=false` tells xTranslator to use the explicit file path instead of\n  the game Data folder.\n- `Command=LoadFile:\u003Cplugin>` loads one `.esm\u002F.esp\u002F.esl`.\n- `Command=ImportSst:\u003CapplyOn>:\u003Cmode>:\u003Csst>` imports an SST dictionary.\n- `Command=Finalize` writes the final output for the loaded file.\n- `Command=CloseFile` closes it before the next rule.\n\nUseful `ImportSst` parameters:\n\n- `applyOn=0`: apply to everything.\n- `applyOn=1`: apply only to not-translated items.\n- `mode=1`: strict FormID + original string matching.\n- `mode=3`: strings-only matching.\n\nPreferred pattern: run strict first, then optionally use strings-only only for\nstill-untranslated items:\n\n```text\nCommand=ImportSst:0:1:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=ImportSst:1:3:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\n```\n\nIf the user is debugging record identity, omit the strings-only fallback so\nunmatched records remain visible.\n\n### Case A - plugin-finalized output\n\nUse this for mods where xTranslator saves the plugin itself, such as a compact\nESM\u002FESP\u002FESL workflow. Do not load the source plugin. First copy the source\nplugin to the sibling MO2 overlay directory named `Original Mod Name - SC`, then\nload that copied file in Batch Processor.\n\n```powershell\n# Prepare the sibling MO2 overlay so xTranslator finalizes a copy, not the source plugin.\nNew-Item -ItemType Directory -Force -Path \"D:\\path\\to\\Mod Name - SC\" | Out-Null\nCopy-Item -LiteralPath \"D:\\path\\to\\Mod Name\\mod.esm\" -Destination \"D:\\path\\to\\Mod Name - SC\\mod.esm\" -Force\n```\n\nThen provide the human this Batch Processor block:\n\n```text\nStartRule\nLangSource=en\nLangDest=zhhans\nUseDataDir=false\nCommand=LoadFile:D:\\path\\to\\Mod Name - SC\\mod.esm\nCommand=ImportSst:0:1:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=ImportSst:1:3:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=Finalize\nCommand=CloseFile\nEndRule\n```\n\nExpected success log contains `Saving Esp\u002FEsm\u002FEsl files...` and a saved path\nunder `Mod Name - SC`.\n\n### Case B - loose Strings output\n\nUse this for string-backed Starfield plugins where xTranslator finalizes\n`strings\\*_zhhans.strings`, `*.dlstrings`, and `*.ilstrings` instead of changing\nthe BA2. Preferred: copy the plugin and text BA2 to `- SC`, load the copied\nplugin there, and let xTranslator write loose strings directly under `- SC`.\n\n```powershell\n# Prepare a sibling overlay with the plugin and the BA2 that contains source strings.\nNew-Item -ItemType Directory -Force -Path \"D:\\path\\to\\Mod Name - SC\" | Out-Null\nCopy-Item -LiteralPath \"D:\\path\\to\\Mod Name\\mod.esm\" -Destination \"D:\\path\\to\\Mod Name - SC\\mod.esm\" -Force\nCopy-Item -LiteralPath \"D:\\path\\to\\Mod Name\\mod - main.ba2\" -Destination \"D:\\path\\to\\Mod Name - SC\\mod - main.ba2\" -Force\n```\n\nBatch Processor block:\n\n```text\nStartRule\nLangSource=en\nLangDest=zhhans\nUseDataDir=false\nCommand=LoadFile:D:\\path\\to\\Mod Name - SC\\mod.esm\nCommand=ImportSst:0:1:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=ImportSst:1:3:C:\\path\\to\\project\\exports\\mod_english_chinese.sst\nCommand=Finalize\nCommand=CloseFile\nEndRule\n```\n\nExpected success log contains `Saving Strings files...` and saved files under\n`Mod Name - SC\\strings\\`.\n\nAfter verifying the target-language loose files exist under `Mod Name -\nSC\\strings\\`, delete the temporary copied plugin\u002Farchive files from `Mod Name -\nSC`. For string-localized mods, the delivered MO2 overlay only needs the loose\n`strings\\*_zhhans.strings`, `*.dlstrings`, and `*.ilstrings` files; keeping the\ncopied `.esm\u002F.esp\u002F.esl`, `.ba2`, or `.bsa` in `- SC` can shadow the source mod\nunnecessarily.\n\n```powershell\n# After successful string finalization, keep only the loose translated strings.\nRemove-Item -LiteralPath \"D:\\path\\to\\Mod Name - SC\\mod.esm\" -Force\nRemove-Item -LiteralPath \"D:\\path\\to\\Mod Name - SC\\mod - main.ba2\" -Force\n```\n\nIf the BA2 is too large to copy and the user accepts a staging step, load the\nsource plugin and let xTranslator write loose target-language strings into the\nsource mod's `strings` folder. Immediately move only the generated target\nlanguage files to `- SC\\strings`:\n\n```powershell\n# Move the loose target-language strings produced by xTranslator into the sibling overlay.\nNew-Item -ItemType Directory -Force -Path \"D:\\path\\to\\Mod Name - SC\\strings\" | Out-Null\nMove-Item -LiteralPath \"D:\\path\\to\\Mod Name\\strings\\mod_zhhans.strings\" -Destination \"D:\\path\\to\\Mod Name - SC\\strings\\mod_zhhans.strings\" -Force\nMove-Item -LiteralPath \"D:\\path\\to\\Mod Name\\strings\\mod_zhhans.dlstrings\" -Destination \"D:\\path\\to\\Mod Name - SC\\strings\\mod_zhhans.dlstrings\" -Force\nMove-Item -LiteralPath \"D:\\path\\to\\Mod Name\\strings\\mod_zhhans.ilstrings\" -Destination \"D:\\path\\to\\Mod Name - SC\\strings\\mod_zhhans.ilstrings\" -Force\n```\n\nOnly use this fallback when the user explicitly accepts temporary writes to the\nsource mod directory. Never write into the game Stock\u002FData directory.\n\n## Browser GUI Flow\n\nUse the GUI when a human needs to configure, inspect, approve, or manually fix:\n\n```powershell\n# Launch the local browser control panel.\nxtl gui\n```\n\nIf the GUI or its process state becomes inconsistent, restart it with the stable\nhelper:\n\n```powershell\n# From tools\u002Fbgs-translator, restart the web GUI on the usual local port.\npowershell -ExecutionPolicy Bypass -File .\\scripts\\restart-web-gui.ps1 -Port 7847\n```\n\nThe GUI is browser-only. There is no Tk fallback.\n\nHigh-signal tabs:\n\n- Project: import plugin projects, see project metadata, export SST, open export\n  directory, and see record signature translation stats.\n- Entries: filter, inspect source\u002Fdestination, quick translate, multi-select,\n  submit selected entries, or use fanatic mode for all filtered entries.\n- Prompt: inspect current pending prompt and approve\u002Fapprove-all\u002Fskip\u002Fdiscard.\n- Batches: watch run progress, request stop, inspect historical audit records,\n  and discard completed run translations when supported.\n- Profiles: create\u002Factivate\u002Fprobe providers and store API keys locally.\n- Glossary: edit player terms and do-not-translate terms.\n- Logs: inspect run files and technical logs.\n\nFor detailed player-facing guidance, point the user to the manuals listed at the\ntop of this skill.\n\nFor maintaining Starfield official terminology packs, mod-specific glossary\npacks, or third-party Skyrim\u002FFallout localization KBs, switch to\n`maintaining-modding-environments`. Translator runs consume those packs; they do\nnot own KB release or provenance policy.\n\n## Prompt Context Requirements\n\nDo not send generic prompts. For every real batch, build context with:\n\n- the game, era, faction\u002Fsetting, and localization style;\n- the mod's purpose, systems, characters, places, terminology, and UI tone;\n- signature explanations only for signatures present in the batch;\n- player glossary and do-not-translate rules;\n- protected placeholder rules such as `{P0}` and `\u003CAlias=...>`.\n\nIf the user provides a mod URL, browse it and summarize relevant features. If\nthe mod page is unavailable, use local readmes\u002Fproject entries and state the\ngap.\n\n## Status Semantics\n\nAlign user-facing status with xTranslator concepts:\n\n- `untranslated`: no accepted translation exists.\n- `translated`: validated\u002Faccepted translation.\n- `needs_review`: partial or suspicious translation that requires human review.\n- `locked`: should remain unchanged and should not be sent to AI.\n\nPlaceholder-only and number-only strings should not enter AI batches. Do not\nhide a mismatch between GUI counts and exported SST; investigate before\nclaiming success.\n\n## Anti-Patterns\n\n- Do not copy examples blindly. Check `--help` for supported values first.\n- Do not make up hard caps. Respect user batch size\u002Fbudget settings; if a prompt\n  may exceed context, report it and ask for smaller batches\u002Fsettings.\n- Do not mark English source copied to destination as translated unless it is a\n  deliberate locked\u002Fno-translate entry.\n- Do not rely on string-only SST matching as proof of correct export. FormID and\n  record identity should match xTranslator where possible.\n- Do not leave orphaned pending-preview state after a service restart. Historical\n  plans are audit records unless a real resume command exists.\n- Do not promise cancellation is free; already-started provider calls may bill.\n",{"name":5,"description":7},"main","b0f0a87593c578a7898eccaf1b363f7f87c1f99b","plugins\u002FBB-84C\u002Fbgs-modding-superpowers\u002Fskills\u002Fusing-bgs-translator",491,"https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins","https:\u002F\u002Fhol.org\u002Fplugins\u002Fbest-codex-plugins","https:\u002F\u002Fgithub.com\u002Fhashgraph-online","https:\u002F\u002Fhol.org","Open-source standards and developer tools for interoperable AI agents: discovery, identity, communication, payments, privacy and security.","2026-11-14 14:06:13",{"source":8,"skillName":5,"commands":61,"prompt":71,"zipAvailable":72},[62,65,68],{"tool":63,"command":64},"npx","npx skills add hashgraph-online\u002Fawesome-codex-plugins --skill using-bgs-translator",{"tool":66,"command":67},"bunx","bunx skills add hashgraph-online\u002Fawesome-codex-plugins --skill using-bgs-translator",{"tool":69,"command":70},"pnpm","pnpm dlx skills add hashgraph-online\u002Fawesome-codex-plugins --skill using-bgs-translator","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002FBB-84C\u002Fbgs-modding-superpowers\u002Fskills\u002Fusing-bgs-translator\u002FSKILL.md",true,[],[75,85,93,101,109,117],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":79,"licenseFileUrl":80,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":84},"01M7DV332YQMGD8WTRJ9DZQXTB","worktable-review-with-annotations","Create, inspect, reply to, update, or resolve situated Worktable annotations. Use when feedback belongs to a Doc section, block, text range, or HTML Doc, or when the user asks to address existing review feedback and coordinate a requested artifact edit.","MIT","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-review-with-annotations\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-review-with-annotations\u002FSKILL.md",[],[],"2026-11-15 14:19:17",{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":79,"licenseFileUrl":89,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":84},"01M7DV2YGHCZVBEB1AHYB6CGHE","worktable-find-and-synthesize-context","Find, assess, and synthesize context from Worktable Docs and Records. Use when the user asks what a workspace contains, wants workspace-backed research, needs current facts reconciled across artifacts, or requests a cited synthesis without changing workspace state.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-find-and-synthesize-context\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-find-and-synthesize-context\u002FSKILL.md",[],[],{"skillNo":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":79,"licenseFileUrl":97,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":84},"01M7DV2TQN1H11ME2SAF9MXN7V","worktable-create-or-update-docs","Create, edit, rename, or maintain narrative Worktable Docs. Use for plans, decisions, briefs, notes, research syntheses, and other durable prose when the user requests a new artifact or a safe update to an existing one.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-update-docs\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-update-docs\u002FSKILL.md",[],[],{"skillNo":102,"name":103,"nameZh":6,"description":104,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":79,"licenseFileUrl":105,"sourceHtmlUrl":106,"categories":107,"keywords":108,"publishedAt":84},"01M7DV2NE452NEA8TXATKSNAJH","worktable-create-or-manage-records","Organize repeated structured information as durable Worktable Records. Use for research datasets, inventories, directories, pipelines, trackers, and other collections whose items need identity, fields, filtering, links, validation, or individual updates.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-records\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-records\u002FSKILL.md",[],[],{"skillNo":110,"name":111,"nameZh":6,"description":112,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":79,"licenseFileUrl":113,"sourceHtmlUrl":114,"categories":115,"keywords":116,"publishedAt":84},"01M7DV2HENRBR2KSCVM064WH0Y","worktable-create-or-manage-interactive-html","Create or improve durable visual and interactive artifacts in Worktable. Use for dashboards, trackers, calculators, mini apps, slide presentations, and other requests that benefit from a polished HTML Doc.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-interactive-html\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-create-or-manage-interactive-html\u002FSKILL.md",[],[],{"skillNo":118,"name":119,"nameZh":6,"description":120,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":79,"licenseFileUrl":121,"sourceHtmlUrl":122,"categories":123,"keywords":124,"publishedAt":84},"01M7DV2CADR5R7RQFFYPNAGR9Y","worktable-collaborate-threads","Read, answer, and hand off messages through durable Worktable threads with users or other authenticated participants. Use when the user asks to check addressed messages, reply to a participant, continue a thread, request another agent's help, wait briefly for a response, or watch for comments and messages sent to the agent.","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-collaborate-threads\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fhashgraph-online\u002Fawesome-codex-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fworktable\u002Fworktable-plugin\u002Fskills\u002Fworktable-collaborate-threads\u002FSKILL.md",[],[]]