[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7EXVQQTDTTQEAX2YY73RF7H":3,"skill-same:01M7EXVQQTDTTQEAX2YY73RF7H":80},{"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":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":27,"commitSha":28,"skillDirectory":29,"forksCount":30,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":33,"ownerDescription":6,"updatedAt":34,"install":35,"resources":47},"01M7EXVQQTDTTQEAX2YY73RF7H","review-and-simplify",null,"Review the current branch or a GitHub PR with risk-tiered specialist subagents, judge their findings, apply safe simplifications to your own branch, and post to the PR on request.","ravnhq\u002Fravn-core","ravnhq","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F32917297?v=4","01M7EXRETGEZ3CYCQTXRQ83BTJ","Ravn","01M7EXREV559052BB2B6AX8X5K",2,"MIT","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: review-and-simplify\ndescription: Review the current branch or a GitHub PR with risk-tiered specialist subagents, judge their findings, apply safe simplifications to your own branch, and post to the PR on request.\ndisable-model-invocation: true\nmetadata:\n  proof: subjective\n---\n\n# Review and Simplify\n\nAdapted from Ravn's Symphony pull request reviewer and Cloudflare's write-up of its AI code reviewer ([\"Cutting our inference bill by 68%\"](https:\u002F\u002Fblog.cloudflare.com\u002Fobsessive-inference-optimization\u002F)).\n\nYou are the **coordinator**. A script picks the change's risk tier. Specialists run in parallel in your own harness and model family, per [models.md](references\u002Fmodels.md), each reviewing one specialty against the same inputs. You judge their findings, apply the safe simplifications, and report one verdict.\n\nResolve `\u003Cskill>` to this installed skill's directory. Run every command from the repo root.\n\n**Scope.** With no argument, review the current branch: everything from its merge base to the working tree, uncommitted and untracked files included. With a PR number or URL, review that PR on the engineer's existing `gh` login. \"Report only\" or `--no-fix` keeps every finding report-only. `--tier \u003Ctrivial|lite|full>` forces a tier; `--full` forces a full first review instead of a re-review; `--max-findings \u003Cn|all>` changes the report's cap of 20 open findings.\n\n**When a script command fails**, it prints `{\"ok\": false, \"code\", \"message\", \"hint\"}`. Correct the cause the code and hint name and run it again. After two failed corrections, do that step by hand under the same contract and name it as done by hand in your report.\n\n## Step 1. Gather the change\n\nMake a temporary directory for this run, `\u003Crun>`. Every diff below uses `\u003Cdiff-flags>`: `--no-color --no-ext-diff -M --src-prefix=a\u002F --dst-prefix=b\u002F`, which fix the output whatever the user's git config says.\n\n**Branch.** Take the base from the user, or else the default branch from `git symbolic-ref --short refs\u002Fremotes\u002Forigin\u002FHEAD`. Fetch it, then run this as one shell call:\n\n```bash\nmb=$(git merge-base \u003Cbase> HEAD)\ngit diff \u003Cdiff-flags> \"$mb\" > \u003Crun>\u002Fdiff.patch\ngit ls-files -z --others --exclude-standard |\n  xargs -0 -I{} git diff \u003Cdiff-flags> --no-index -- \u002Fdev\u002Fnull {} >> \u003Crun>\u002Fdiff.patch   # exit 1 per file is normal\ngit status --porcelain                                      # the user's uncommitted files; keep for Step 5\n```\n\nThe head checkout is the repo root, and `\u003Chead>` is `git rev-parse HEAD`.\n\n**Pull request.** Read `gh pr view \u003Cpr> --json number,url,headRefOid,baseRefName,author`. Fetch with `git fetch origin \u003CbaseRefName> pull\u002F\u003Cpr>\u002Fhead`. Then `\u003Cmerge-base>` is `git merge-base origin\u002F\u003CbaseRefName> \u003CheadRefOid>` and `\u003Chead>` is `headRefOid`. Write `git diff \u003Cdiff-flags> \u003Cmerge-base> \u003Chead> > \u003Crun>\u002Fdiff.patch`. When `HEAD` is `\u003Chead>` and `git status --porcelain` is empty, the head checkout is the repo root. Otherwise create one with `git -c core.symlinks=false worktree add --detach \u003Crun>\u002Fhead \u003Chead>`, so a symlink in the change cannot point a reader outside it.\n\n**Guidance.** Write `\u003Crun>\u002Fguidance.md` from the merge base, so the guidance the change is judged by comes from before it: every `AGENTS.md`, `CLAUDE.md`, `REVIEW.md`, and `.claude\u002Frules\u002F*` at the root or in a directory that holds a changed file. List them with `git ls-tree -r --name-only \u003Cmerge-base>`, then write each, in one shell loop, as a `## \u003Cpath>` heading followed by `git show \u003Cmerge-base>:\u003Cpath>`. Use an empty file when none exist.\n\n**Re-review.** The state file is `$(git rev-parse --git-common-dir)\u002Freview-and-simplify\u002F\u003Ckey>.json`, with `\u003Ckey>` as the branch name or `pr-\u003Cnumber>`, `\u002F` replaced by `-`. When it exists, the user did not pass `--full`, its `merge_base` equals `\u003Cmerge-base>`, and `git merge-base --is-ancestor \u003Csaved head> \u003Chead>` succeeds, this run is a re-review. Write the changes since then to `\u003Crun>\u002Fincrement.patch`: `git diff \u003Cdiff-flags> \u003Csaved head>` with untracked files appended for a branch, or `git diff \u003Cdiff-flags> \u003Csaved head> \u003Chead>` for a pull request. Specialists review the increment, and the state's `findings` are the previous findings. An empty increment means nothing changed: say so and stop, unless the user asks for `--full`.\n\nDone when `\u003Crun>\u002Fdiff.patch` holds the whole change and `\u003Crun>\u002Fguidance.md` exists.\n\n## Step 2. Pick the tier\n\n```bash\nnode \u003Cskill>\u002Fscripts\u002Freview.ts tier \u003Crun>\u002Fincrement.patch   # re-review; otherwise \u003Crun>\u002Fdiff.patch\n```\n\nAdd `--tier` when the user forced one. Save the JSON as `\u003Crun>\u002Ftier.json`: the `tier` and its reason, the `specialists`, each with a `model` class, and the `files` they review.\n\nDrop a listed specialist that has nothing to review: `release` when no changed file is a build, dependency, CI, container, migration, or runtime-configuration file, and `engineering_standards` when `guidance.md` is empty. On a re-review, also drop one whose specialty the increment gives nothing to check and that has no open previous finding. `security` always runs when listed.\n\nDone when `\u003Crun>\u002Ftier.json` exists and the dropped specialists are named.\n\n## Step 3. Run the specialists\n\nGive every specialist the same brief, identical up to its last line, so the requests share a cached prefix:\n\n```text\nYou are a review-and-simplify specialist. Before anything else, read these three files in full, in one parallel batch, then follow the first one:\n- \u003Cskill>\u002Freferences\u002Freviewer.md\n- \u003Crun>\u002Fguidance.md\n- \u003Crun>\u002Fdiff.patch\nHead checkout: \u003Ccheckout>\nSpecialty: \u003Cname>\n```\n\nOn a re-review, the third file is `\u003Crun>\u002Fincrement.patch`.\n\nOn the `trivial` tier, run the one `code_quality` pass yourself from the same brief, since a change that small does not repay a subagent. Write its findings JSON down as a specialist's reply; Steps 4 to 7 still apply.\n\nOn the other tiers, spawn each specialist as a fresh read-only subagent, not a fork of your conversation, on the model its class names in [models.md](references\u002Fmodels.md):\n\n- **Claude Code:** the Agent tool with ravn-core's read-only `review-specialist` agent: `ravn-core:review-specialist` from the plugin, or `review-specialist` where it is installed as a project agent.\n- **Codex:** `spawn_agent` with the brief as its message.\n- **A harness without subagents:** read the three files once, then run each specialist's pass yourself in turn, changing only the specialty, and write its reply down before you start the next.\n\nSpawn the first specialist of each model class, then the rest together in your next message without waiting, so the later requests reuse the prefix the first ones cache. Wait for every reply. Harnesses cap concurrent subagents (Codex's `agents.max_concurrent_threads_per_session`, for one): when a spawn is refused, retry it as soon as a running one replies, and when none will start, run its pass yourself as a harness without subagents does. A reply that is not the findings JSON gets one follow-up asking for only the JSON.\n\nDone when you hold a valid findings reply, from a subagent or your own pass, for every specialist you did not drop in Step 2. Report a specialist as run only when you hold its reply; one without a reply is failed.\n\n## Step 4. Judge\n\nJudge the combined findings into `\u003Crun>\u002Fjudged.json`, as `{\"findings\": [...]}`. Each finding carries `specialist` plus the fields in [reviewer.md](references\u002Freviewer.md#output):\n\n- Keep an issue that several specialists raised at one location once, under the specialty that fits it best, and list the others in `also_raised_by`.\n- Move a finding filed under the wrong specialty to the right one.\n- Set each severity from the impact the finding demonstrates, whatever its specialist chose. A `suggestion` that shows a concrete security or correctness impact, such as a token written to logs, rises to that impact.\n- Drop speculative findings, nitpicks, false positives, and findings that contradict `guidance.md`; read the code to settle a doubt.\n- Keep a `replacement` only when it is exact code for the anchored lines.\n- On a re-review, read the code at each previous finding's anchor. Drop the ones the code now fixes. Re-emit the ones still unfixed, at their anchor in the current diff. Re-emit one whose PR thread a person resolved only when the issue has grown worse, at the higher severity.\n\nDone when every specialist finding is kept, merged, moved, or dropped for a reason you can state.\n\n## Step 5. Apply simplifications\n\nWhen `simplify` ran and you kept any of its findings, follow [simplify.md](references\u002Fsimplify.md). It decides whether they apply. When they do, it applies them, proves the repo's checks still pass, and marks each one `applied` or `reverted` in `\u003Crun>\u002Fjudged.json`.\n\n## Step 6. Check\n\n```bash\nnode \u003Cskill>\u002Fscripts\u002Freview.ts check \u003Crun>\u002Fdiff.patch \u003Crun>\u002Fjudged.json --commit \u003Chead>\n```\n\nAdd `--max-findings \u003Cn>` or `--max-findings all` when the user asked for a different cap.\n\nCheck against `\u003Crun>\u002Fdiff.patch` even on a re-review: it is the diff GitHub shows. The script rejects and names each finding anchored outside that diff's hunks, drops exact duplicates, caps the open findings, computes the verdict from the unapplied ones, and returns the `report` Markdown. It writes `report.md` and `github-review.json` beside `judged.json`.\n\nDone when it exits 0.\n\n## Step 7. Report\n\nSave `{\"head\": \"\u003Chead>\", \"merge_base\": \"\u003Cmerge-base>\", \"findings\": [...]}` to the state file, keeping only findings whose status is not `applied`.\n\nReply with the `report` from Step 6 as it is, verdict line first, then the tier and why, the specialists that ran, were skipped, or failed, and for applied simplifications their `git diff --stat` and the checks that passed.\n\nPost to the pull request only when the user asks:\n\n- **Pull request:** `gh api --method POST repos\u002F\u003Cowner>\u002F\u003Crepo>\u002Fpulls\u002F\u003Cnumber>\u002Freviews --input \u003Crun>\u002Fgithub-review.json`.\n- **Branch:** first find its pull request with `gh pr view --json number,url,headRefOid`. Post only when `headRefOid` is `\u003Chead>` and Step 1's `git status --porcelain` was empty, so every anchor exists on GitHub. Otherwise tell the user to push and review again.\n\nThe review's event is always `COMMENT`: the verdict is text, and the review never approves or blocks. Report the posted review's `html_url`.\n\nFinish by removing `\u003Crun>`, and `git worktree remove --force \u003Crun>\u002Fhead` when you created it.\n","# Review and Simplify\n\nAdapted from Ravn's Symphony pull request reviewer and Cloudflare's write-up of its AI code reviewer ([\"Cutting our inference bill by 68%\"](https:\u002F\u002Fblog.cloudflare.com\u002Fobsessive-inference-optimization\u002F)).\n\nYou are the **coordinator**. A script picks the change's risk tier. Specialists run in parallel in your own harness and model family, per [models.md](references\u002Fmodels.md), each reviewing one specialty against the same inputs. You judge their findings, apply the safe simplifications, and report one verdict.\n\nResolve `\u003Cskill>` to this installed skill's directory. Run every command from the repo root.\n\n**Scope.** With no argument, review the current branch: everything from its merge base to the working tree, uncommitted and untracked files included. With a PR number or URL, review that PR on the engineer's existing `gh` login. \"Report only\" or `--no-fix` keeps every finding report-only. `--tier \u003Ctrivial|lite|full>` forces a tier; `--full` forces a full first review instead of a re-review; `--max-findings \u003Cn|all>` changes the report's cap of 20 open findings.\n\n**When a script command fails**, it prints `{\"ok\": false, \"code\", \"message\", \"hint\"}`. Correct the cause the code and hint name and run it again. After two failed corrections, do that step by hand under the same contract and name it as done by hand in your report.\n\n## Step 1. Gather the change\n\nMake a temporary directory for this run, `\u003Crun>`. Every diff below uses `\u003Cdiff-flags>`: `--no-color --no-ext-diff -M --src-prefix=a\u002F --dst-prefix=b\u002F`, which fix the output whatever the user's git config says.\n\n**Branch.** Take the base from the user, or else the default branch from `git symbolic-ref --short refs\u002Fremotes\u002Forigin\u002FHEAD`. Fetch it, then run this as one shell call:\n\n```bash\nmb=$(git merge-base \u003Cbase> HEAD)\ngit diff \u003Cdiff-flags> \"$mb\" > \u003Crun>\u002Fdiff.patch\ngit ls-files -z --others --exclude-standard |\n  xargs -0 -I{} git diff \u003Cdiff-flags> --no-index -- \u002Fdev\u002Fnull {} >> \u003Crun>\u002Fdiff.patch   # exit 1 per file is normal\ngit status --porcelain                                      # the user's uncommitted files; keep for Step 5\n```\n\nThe head checkout is the repo root, and `\u003Chead>` is `git rev-parse HEAD`.\n\n**Pull request.** Read `gh pr view \u003Cpr> --json number,url,headRefOid,baseRefName,author`. Fetch with `git fetch origin \u003CbaseRefName> pull\u002F\u003Cpr>\u002Fhead`. Then `\u003Cmerge-base>` is `git merge-base origin\u002F\u003CbaseRefName> \u003CheadRefOid>` and `\u003Chead>` is `headRefOid`. Write `git diff \u003Cdiff-flags> \u003Cmerge-base> \u003Chead> > \u003Crun>\u002Fdiff.patch`. When `HEAD` is `\u003Chead>` and `git status --porcelain` is empty, the head checkout is the repo root. Otherwise create one with `git -c core.symlinks=false worktree add --detach \u003Crun>\u002Fhead \u003Chead>`, so a symlink in the change cannot point a reader outside it.\n\n**Guidance.** Write `\u003Crun>\u002Fguidance.md` from the merge base, so the guidance the change is judged by comes from before it: every `AGENTS.md`, `CLAUDE.md`, `REVIEW.md`, and `.claude\u002Frules\u002F*` at the root or in a directory that holds a changed file. List them with `git ls-tree -r --name-only \u003Cmerge-base>`, then write each, in one shell loop, as a `## \u003Cpath>` heading followed by `git show \u003Cmerge-base>:\u003Cpath>`. Use an empty file when none exist.\n\n**Re-review.** The state file is `$(git rev-parse --git-common-dir)\u002Freview-and-simplify\u002F\u003Ckey>.json`, with `\u003Ckey>` as the branch name or `pr-\u003Cnumber>`, `\u002F` replaced by `-`. When it exists, the user did not pass `--full`, its `merge_base` equals `\u003Cmerge-base>`, and `git merge-base --is-ancestor \u003Csaved head> \u003Chead>` succeeds, this run is a re-review. Write the changes since then to `\u003Crun>\u002Fincrement.patch`: `git diff \u003Cdiff-flags> \u003Csaved head>` with untracked files appended for a branch, or `git diff \u003Cdiff-flags> \u003Csaved head> \u003Chead>` for a pull request. Specialists review the increment, and the state's `findings` are the previous findings. An empty increment means nothing changed: say so and stop, unless the user asks for `--full`.\n\nDone when `\u003Crun>\u002Fdiff.patch` holds the whole change and `\u003Crun>\u002Fguidance.md` exists.\n\n## Step 2. Pick the tier\n\n```bash\nnode \u003Cskill>\u002Fscripts\u002Freview.ts tier \u003Crun>\u002Fincrement.patch   # re-review; otherwise \u003Crun>\u002Fdiff.patch\n```\n\nAdd `--tier` when the user forced one. Save the JSON as `\u003Crun>\u002Ftier.json`: the `tier` and its reason, the `specialists`, each with a `model` class, and the `files` they review.\n\nDrop a listed specialist that has nothing to review: `release` when no changed file is a build, dependency, CI, container, migration, or runtime-configuration file, and `engineering_standards` when `guidance.md` is empty. On a re-review, also drop one whose specialty the increment gives nothing to check and that has no open previous finding. `security` always runs when listed.\n\nDone when `\u003Crun>\u002Ftier.json` exists and the dropped specialists are named.\n\n## Step 3. Run the specialists\n\nGive every specialist the same brief, identical up to its last line, so the requests share a cached prefix:\n\n```text\nYou are a review-and-simplify specialist. Before anything else, read these three files in full, in one parallel batch, then follow the first one:\n- \u003Cskill>\u002Freferences\u002Freviewer.md\n- \u003Crun>\u002Fguidance.md\n- \u003Crun>\u002Fdiff.patch\nHead checkout: \u003Ccheckout>\nSpecialty: \u003Cname>\n```\n\nOn a re-review, the third file is `\u003Crun>\u002Fincrement.patch`.\n\nOn the `trivial` tier, run the one `code_quality` pass yourself from the same brief, since a change that small does not repay a subagent. Write its findings JSON down as a specialist's reply; Steps 4 to 7 still apply.\n\nOn the other tiers, spawn each specialist as a fresh read-only subagent, not a fork of your conversation, on the model its class names in [models.md](references\u002Fmodels.md):\n\n- **Claude Code:** the Agent tool with ravn-core's read-only `review-specialist` agent: `ravn-core:review-specialist` from the plugin, or `review-specialist` where it is installed as a project agent.\n- **Codex:** `spawn_agent` with the brief as its message.\n- **A harness without subagents:** read the three files once, then run each specialist's pass yourself in turn, changing only the specialty, and write its reply down before you start the next.\n\nSpawn the first specialist of each model class, then the rest together in your next message without waiting, so the later requests reuse the prefix the first ones cache. Wait for every reply. Harnesses cap concurrent subagents (Codex's `agents.max_concurrent_threads_per_session`, for one): when a spawn is refused, retry it as soon as a running one replies, and when none will start, run its pass yourself as a harness without subagents does. A reply that is not the findings JSON gets one follow-up asking for only the JSON.\n\nDone when you hold a valid findings reply, from a subagent or your own pass, for every specialist you did not drop in Step 2. Report a specialist as run only when you hold its reply; one without a reply is failed.\n\n## Step 4. Judge\n\nJudge the combined findings into `\u003Crun>\u002Fjudged.json`, as `{\"findings\": [...]}`. Each finding carries `specialist` plus the fields in [reviewer.md](references\u002Freviewer.md#output):\n\n- Keep an issue that several specialists raised at one location once, under the specialty that fits it best, and list the others in `also_raised_by`.\n- Move a finding filed under the wrong specialty to the right one.\n- Set each severity from the impact the finding demonstrates, whatever its specialist chose. A `suggestion` that shows a concrete security or correctness impact, such as a token written to logs, rises to that impact.\n- Drop speculative findings, nitpicks, false positives, and findings that contradict `guidance.md`; read the code to settle a doubt.\n- Keep a `replacement` only when it is exact code for the anchored lines.\n- On a re-review, read the code at each previous finding's anchor. Drop the ones the code now fixes. Re-emit the ones still unfixed, at their anchor in the current diff. Re-emit one whose PR thread a person resolved only when the issue has grown worse, at the higher severity.\n\nDone when every specialist finding is kept, merged, moved, or dropped for a reason you can state.\n\n## Step 5. Apply simplifications\n\nWhen `simplify` ran and you kept any of its findings, follow [simplify.md](references\u002Fsimplify.md). It decides whether they apply. When they do, it applies them, proves the repo's checks still pass, and marks each one `applied` or `reverted` in `\u003Crun>\u002Fjudged.json`.\n\n## Step 6. Check\n\n```bash\nnode \u003Cskill>\u002Fscripts\u002Freview.ts check \u003Crun>\u002Fdiff.patch \u003Crun>\u002Fjudged.json --commit \u003Chead>\n```\n\nAdd `--max-findings \u003Cn>` or `--max-findings all` when the user asked for a different cap.\n\nCheck against `\u003Crun>\u002Fdiff.patch` even on a re-review: it is the diff GitHub shows. The script rejects and names each finding anchored outside that diff's hunks, drops exact duplicates, caps the open findings, computes the verdict from the unapplied ones, and returns the `report` Markdown. It writes `report.md` and `github-review.json` beside `judged.json`.\n\nDone when it exits 0.\n\n## Step 7. Report\n\nSave `{\"head\": \"\u003Chead>\", \"merge_base\": \"\u003Cmerge-base>\", \"findings\": [...]}` to the state file, keeping only findings whose status is not `applied`.\n\nReply with the `report` from Step 6 as it is, verdict line first, then the tier and why, the specialists that ran, were skipped, or failed, and for applied simplifications their `git diff --stat` and the checks that passed.\n\nPost to the pull request only when the user asks:\n\n- **Pull request:** `gh api --method POST repos\u002F\u003Cowner>\u002F\u003Crepo>\u002Fpulls\u002F\u003Cnumber>\u002Freviews --input \u003Crun>\u002Fgithub-review.json`.\n- **Branch:** first find its pull request with `gh pr view --json number,url,headRefOid`. Post only when `headRefOid` is `\u003Chead>` and Step 1's `git status --porcelain` was empty, so every anchor exists on GitHub. Otherwise tell the user to push and review again.\n\nThe review's event is always `COMMENT`: the verdict is text, and the review never approves or blocks. Report the posted review's `html_url`.\n\nFinish by removing `\u003Crun>`, and `git worktree remove --force \u003Crun>\u002Fhead` when you created it.\n",{"name":5,"metadata":24,"description":7,"disable-model-invocation":26},{"proof":25},"subjective",true,"main","8debc762560573209acaf70a1a66f1061d7c8060","plugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify",0,"https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core","https:\u002F\u002Fgithub.com\u002Fravnhq","https:\u002F\u002Fwww.ravn.co","2026-11-13 05:50:45",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":26},[37,40,43],{"tool":38,"command":39},"npx","npx skills add ravnhq\u002Fravn-core --skill review-and-simplify",{"tool":41,"command":42},"bunx","bunx skills add ravnhq\u002Fravn-core --skill review-and-simplify",{"tool":44,"command":45},"pnpm","pnpm dlx skills add ravnhq\u002Fravn-core --skill review-and-simplify","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002FSKILL.md",[48,54,58,62,66,72,76],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"agents\u002Fopenai.yaml","REFERENCE","参考文档",193,"https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002Fagents\u002Fopenai.yaml",{"path":55,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Fmodels.md",1046,"https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002Freferences\u002Fmodels.md",{"path":59,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Freviewer.md",7244,"https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002Freferences\u002Freviewer.md",{"path":63,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Fsimplify.md",1983,"https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002Freferences\u002Fsimplify.md",{"path":67,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":70,"sourceHtmlUrl":71},"scripts\u002Fdiff.ts","SCRIPT","脚本",6102,"https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002Fscripts\u002Fdiff.ts",{"path":73,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":74,"sourceHtmlUrl":75},"scripts\u002Ffindings.ts",11733,"https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002Fscripts\u002Ffindings.ts",{"path":77,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":78,"sourceHtmlUrl":79},"scripts\u002Freview.ts",7929,"https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Freview-and-simplify\u002Fscripts\u002Freview.ts",[81,89,97,105,113,121],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":85,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7EXWEDTSBBB0DQENHSBXCV5","writing-for-agents","Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fwriting-for-agents\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fwriting-for-agents\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":93,"sourceHtmlUrl":94,"categories":95,"keywords":96,"publishedAt":19},"01M7EXWA9QD5H1Z7ZNZHFV7M5E","why","Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed thresholds. Queries every available evidence MCP in parallel and returns a cited read on decisions and tradeoffs. Use how for runtime behavior.","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fwhy\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fwhy\u002FSKILL.md",[],[],{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":101,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":19},"01M7EXW7M17CN7V62E25PM4QAJ","typescript-best-practices","TypeScript type-safety rules: discriminated unions, branded types, no `as`, exhaustive switches, schema-derived types. Use when writing, editing, or reviewing TypeScript.","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Ftypescript-best-practices\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Ftypescript-best-practices\u002FSKILL.md",[],[],{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":109,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":19},"01M7EXW4Q6KTG9RM0R1MB29H9D","show-me-your-work","Keep a reviewable decision trail for unattended or multi-phase work: a TSV log with one row per decision (what, why, evidence, result).","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fshow-me-your-work\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fshow-me-your-work\u002FSKILL.md",[],[],{"skillNo":114,"name":115,"nameZh":6,"description":116,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":117,"sourceHtmlUrl":118,"categories":119,"keywords":120,"publishedAt":19},"01M7EXW00F0AQWA8F8ANCN2ZZT","ship","Take a change from idea to a ready PR with passing CI, running research, grilling, design, verification, and the pr skill in sequence.","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fship\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fship\u002FSKILL.md",[],[],{"skillNo":122,"name":123,"nameZh":6,"description":124,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":125,"sourceHtmlUrl":126,"categories":127,"keywords":128,"publishedAt":19},"01M7EXVTVXP3RB0NPVTPG05YC4","session-pickup","Resume unfinished work from a handoff note or prior coding-agent session. Use before reading a supplied handoff, even when its next action is only an inspection or report. Also use to pick up, continue where you left off, or take over from a transcript, agent URL, or pushed branch. Use pause-safely to write a handoff.","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fsession-pickup\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fravnhq\u002Fravn-core\u002Fblob\u002Fmain\u002Fplugins\u002Fagent\u002Fravn-core\u002Fskills\u002Fsession-pickup\u002FSKILL.md",[],[]]