[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C4XY0AR0VGJ5KKP26F2V67":3,"skill-same:01M7C4XY0AR0VGJ5KKP26F2V67":63},{"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":25,"commitSha":26,"skillDirectory":27,"forksCount":14,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":30,"ownerDescription":6,"updatedAt":31,"install":32,"resources":44},"01M7C4XY0AR0VGJ5KKP26F2V67","deploy-verify",null,"Verify deployed infra after merge\u002Frestart. Triggers: PR merged, deployed, daemon\u002FMCP restart, live probe.","EtanHey\u002Fgolems","EtanHey","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F93338119?v=4","01M7C4SKC9HJC3B9TAX0D9MZ1E","Etan Heyman","01M7C4SKCT59Z49TTN4WWDDEYD",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fdeploy-verify\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: deploy-verify\ndescription: \"Verify deployed infra after merge\u002Frestart. Triggers: PR merged, deployed, daemon\u002FMCP restart, live probe.\"\ndisable-model-invocation: true\n---\n\n# Skill: Deploy Verify\n\n> A merged PR is not a deployed system. Verify the runtime, or do not claim the deploy is done.\n\n## When to Use\n\n- After any infra PR merges\n- When a worker, collab entry, or review comment says \"deployed\", \"restarted\", or \"now running\"\n- Before closing a pane, removing a worktree, or telling the user a runtime fix shipped\n- For LaunchAgents, daemons, CLI tools, MCP servers, and app bundles whose merged code must be reflected in a live artifact\n\nInfra PRs include:\n- LaunchAgent plist changes\n- daemon or socket server changes\n- CLI binary changes\n- MCP server changes\n- app bundle or build-stamp changes\n\n## Composition\n\nThis skill composes directly with:\n\n- `\u002Fnever-fabricate`\n- `\u002Fpost-merge-deploy-check`\n\n`\u002Fpost-merge-deploy-check` proves merged-PR metadata, registry state, bundle provenance, and canonical-path\u002Fprocess drift. This skill is the agent-side workflow that decides when to run that script, what additional runtime checks must be collected, and when \"done\" is still unverified.\n\n## The Rule\n\n**After an infra PR merges, do not claim deployed until all four verification steps pass.**\n\nRequired sequence:\n\n1. process replacement\n2. `launchctl print` status\n3. build-stamp match\n4. end-to-end live probe\n\nIf any step fails, the verdict is `NOT DEPLOYED`. Fix the deploy or open a fix PR. Do not narrate success.\n\n## Step 0: Classify the Merge\n\nBefore running the sequence, classify the PR:\n\n- `no_deploy_required`: docs, skill-only, or other artifact-free changes\n- `deploy_bearing`: LaunchAgent, daemon, CLI, MCP server, app bundle, socket service, or anything that changes live runtime behavior\n\nIf the PR is `no_deploy_required`, say so explicitly and stop.\n\nIf it is `deploy_bearing`, continue through all four steps.\n\n## The Four-Step Sequence\n\n### 1. Process Replacement\n\nGoal: prove the old runtime exited and a fresh runtime is now the one serving requests.\n\nRequired evidence:\n- the old PID is gone, replaced, or otherwise no longer the serving process\n- the new PID exists when the service is expected to be running\n- the running command path does not point at a stale worktree or dev artifact\n\nAcceptable checks:\n- `pgrep -alf \u003Cprocess-pattern>`\n- `ps -p \u003Cpid> -o pid=,ppid=,etime=,command=`\n- path checks against known stale-process patterns from the canonical deploy registry\n\nFailure examples:\n- old PID still alive after a merge that required restart\n- multiple copies of the same app or daemon are serving simultaneously\n- process path still resolves into a worktree or `*-DEV-*` app bundle\n\n### 2. LaunchAgent Status\n\nGoal: prove launchd knows about the correct service state.\n\nRequired evidence:\n- `launchctl print gui\u002F$(id -u)\u002F\u003Cservice-label>` or equivalent for the relevant domain\n- state is consistent with the intended deploy outcome\n- `Program` or `ProgramArguments` point at the canonical artifact, not stale paths\n\nFailure examples:\n- `Could not find service`\n- state still points at the pre-merge path\n- launchd reports the job loaded but the artifact path is wrong\n\n### 3. Build-Stamp Match\n\nGoal: prove the artifact serving requests was built from the merged mainline commit.\n\nRequired evidence:\n- for app bundles: `Info.plist` provenance fields such as `GitCommit`, `BuildTimeUTC`, and expected bundle metadata\n- for CLIs or daemons: `--version`, `version.json`, or equivalent commit\u002Fbuild stamp\n- the stamped commit matches the merged main branch commit or a descendant on canonical main\n\nRequired action:\n- run `\u002Fpost-merge-deploy-check` when the component is in the canonical deploy registry\n\nWhat `\u002Fpost-merge-deploy-check` covers:\n- PR is merged\n- merge target is the canonical branch\n- registry entry resolves to the canonical repo path\n- bundle `GitCommit` matches the registry\n- deployed artifact is at or after the merged PR commit\n- running process path does not match stale forbidden patterns\n\nFailure examples:\n- app version updated but commit stamp still points at the old build\n- merged PR exists but the deployed artifact lineage does not include it\n- registry matches an older build while the collab says \"done\"\n\n### 4. End-to-End Live Probe\n\nGoal: prove the runtime actually serves post-merge behavior, not just the right metadata.\n\nUse the narrowest real probe that exercises the changed path:\n\n- actual MCP tool call\n- socket ping or request\u002Fresponse round-trip\n- CLI invocation against the real installed binary\n- HTTP endpoint hit\n- app-side behavior probe when the change is app-facing\n\nThe probe must hit the changed behavior, not just process liveness.\n\nFailure examples:\n- socket accepts connections but returns old behavior\n- MCP server starts but the changed tool path still fails\n- app bundle stamp is correct but the launched service is not the one responding\n\n## Decision Tree\n\n### 1. Merge report arrives\n\nExample inputs:\n- \"PR #264 merged\"\n- \"worker says the daemon is restarted\"\n- \"collab says deployed\"\n\nAction:\n- classify whether this is `deploy_bearing`\n- if yes, start the four-step sequence\n\n### 2. A worker says `TASK_DONE`\n\nDo not treat this as deployment proof.\n\nAction:\n- require process evidence\n- require `launchctl print`\n- require build-stamp evidence\n- require a live probe\n\nUntil those exist, the status is `UNVERIFIED DEPLOY CLAIM`.\n\n### 3. `\u002Fpost-merge-deploy-check` passes but the live probe fails\n\nVerdict:\n- `NOT DEPLOYED`\n\nReason:\n- provenance is necessary but insufficient; runtime behavior is still wrong\n\n### 4. Live probe passes but build stamp fails\n\nVerdict:\n- `DRIFTED OR NON-CANONICAL DEPLOY`\n\nReason:\n- behavior may look right now, but the agent cannot claim the merged PR is what is serving\n\n## Output Contract\n\nWhen reporting verification, use this structure:\n\n```text\nDeploy verification for \u003Crepo>#\u003Cpr>:\n- Step 1 Process replacement: PASS|FAIL — \u003Cevidence>\n- Step 2 LaunchAgent status: PASS|FAIL — \u003Cevidence>\n- Step 3 Build-stamp match: PASS|FAIL — \u003Cevidence>\n- Step 4 End-to-end probe: PASS|FAIL — \u003Cevidence>\n- Verdict: DEPLOYED | NOT DEPLOYED | NO DEPLOY REQUIRED\n```\n\nIf any evidence is missing, say `UNVERIFIED` and keep the verdict negative.\n\n## Hard Rules\n\n### 1. Merge Is Not Deploy\n\nNever answer \"done\" or \"deployed\" from GitHub merge state alone.\n\n### 2. Subagent Report Is Not Deploy\n\n`TASK_DONE`, collab notes, and \"restarted locally\" claims are not verification. `\u002Fnever-fabricate` applies in full.\n\n### 3. Provenance Alone Is Not Deploy\n\nPassing `\u002Fpost-merge-deploy-check` is not enough without the end-to-end probe.\n\n### 4. Liveness Alone Is Not Deploy\n\nA running PID or loaded LaunchAgent is not enough without build-stamp evidence.\n\n### 5. Clean Up Only After Verification\n\nDo not close the pane, remove the worktree, or report success until all required steps pass.\n\n## Anti-Patterns\n\n| Anti-pattern | Why it fails | Fix |\n|---|---|---|\n| \"PR merged, so deploy is done\" | GitHub state is not runtime state | Run all four verification steps |\n| \"worker said restarted\" | Worker self-report is not proof | Collect process, launchctl, stamp, and probe evidence |\n| \"PID exists, looks good\" | Liveness does not prove the right build or behavior | Check stamp and probe |\n| \"build stamp matches, so done\" | Provenance does not prove serving behavior | Run the live probe |\n| \"probe works, close pane\" | Probe without canonical provenance can hide drift | Run `\u002Fpost-merge-deploy-check` and record the receipt |\n\n## Receipts\n\nRecommended commands include:\n\n```bash\n: \"${ORCHESTRATOR_REPO:?ORCHESTRATOR_REPO must be set}\"\npgrep -alf '\u003Cpattern>'\nlaunchctl print gui\u002F$(id -u)\u002F\u003Cservice-label>\npython3 \"$ORCHESTRATOR_REPO\u002Fscripts\u002Fpost-merge-deploy-check.py\" \u003Crepo> \u003Cpr-number>\n\u003Ccanonical-binary> --version\n\u003Creal socket, MCP, or endpoint probe>\n```\n\nUse the canonical deploy registry when available. If the component is not registered yet, say so explicitly and still complete the process, `launchctl`, build-stamp, and live-probe checks with whatever local evidence is available.\n","# Skill: Deploy Verify\n\n> A merged PR is not a deployed system. Verify the runtime, or do not claim the deploy is done.\n\n## When to Use\n\n- After any infra PR merges\n- When a worker, collab entry, or review comment says \"deployed\", \"restarted\", or \"now running\"\n- Before closing a pane, removing a worktree, or telling the user a runtime fix shipped\n- For LaunchAgents, daemons, CLI tools, MCP servers, and app bundles whose merged code must be reflected in a live artifact\n\nInfra PRs include:\n- LaunchAgent plist changes\n- daemon or socket server changes\n- CLI binary changes\n- MCP server changes\n- app bundle or build-stamp changes\n\n## Composition\n\nThis skill composes directly with:\n\n- `\u002Fnever-fabricate`\n- `\u002Fpost-merge-deploy-check`\n\n`\u002Fpost-merge-deploy-check` proves merged-PR metadata, registry state, bundle provenance, and canonical-path\u002Fprocess drift. This skill is the agent-side workflow that decides when to run that script, what additional runtime checks must be collected, and when \"done\" is still unverified.\n\n## The Rule\n\n**After an infra PR merges, do not claim deployed until all four verification steps pass.**\n\nRequired sequence:\n\n1. process replacement\n2. `launchctl print` status\n3. build-stamp match\n4. end-to-end live probe\n\nIf any step fails, the verdict is `NOT DEPLOYED`. Fix the deploy or open a fix PR. Do not narrate success.\n\n## Step 0: Classify the Merge\n\nBefore running the sequence, classify the PR:\n\n- `no_deploy_required`: docs, skill-only, or other artifact-free changes\n- `deploy_bearing`: LaunchAgent, daemon, CLI, MCP server, app bundle, socket service, or anything that changes live runtime behavior\n\nIf the PR is `no_deploy_required`, say so explicitly and stop.\n\nIf it is `deploy_bearing`, continue through all four steps.\n\n## The Four-Step Sequence\n\n### 1. Process Replacement\n\nGoal: prove the old runtime exited and a fresh runtime is now the one serving requests.\n\nRequired evidence:\n- the old PID is gone, replaced, or otherwise no longer the serving process\n- the new PID exists when the service is expected to be running\n- the running command path does not point at a stale worktree or dev artifact\n\nAcceptable checks:\n- `pgrep -alf \u003Cprocess-pattern>`\n- `ps -p \u003Cpid> -o pid=,ppid=,etime=,command=`\n- path checks against known stale-process patterns from the canonical deploy registry\n\nFailure examples:\n- old PID still alive after a merge that required restart\n- multiple copies of the same app or daemon are serving simultaneously\n- process path still resolves into a worktree or `*-DEV-*` app bundle\n\n### 2. LaunchAgent Status\n\nGoal: prove launchd knows about the correct service state.\n\nRequired evidence:\n- `launchctl print gui\u002F$(id -u)\u002F\u003Cservice-label>` or equivalent for the relevant domain\n- state is consistent with the intended deploy outcome\n- `Program` or `ProgramArguments` point at the canonical artifact, not stale paths\n\nFailure examples:\n- `Could not find service`\n- state still points at the pre-merge path\n- launchd reports the job loaded but the artifact path is wrong\n\n### 3. Build-Stamp Match\n\nGoal: prove the artifact serving requests was built from the merged mainline commit.\n\nRequired evidence:\n- for app bundles: `Info.plist` provenance fields such as `GitCommit`, `BuildTimeUTC`, and expected bundle metadata\n- for CLIs or daemons: `--version`, `version.json`, or equivalent commit\u002Fbuild stamp\n- the stamped commit matches the merged main branch commit or a descendant on canonical main\n\nRequired action:\n- run `\u002Fpost-merge-deploy-check` when the component is in the canonical deploy registry\n\nWhat `\u002Fpost-merge-deploy-check` covers:\n- PR is merged\n- merge target is the canonical branch\n- registry entry resolves to the canonical repo path\n- bundle `GitCommit` matches the registry\n- deployed artifact is at or after the merged PR commit\n- running process path does not match stale forbidden patterns\n\nFailure examples:\n- app version updated but commit stamp still points at the old build\n- merged PR exists but the deployed artifact lineage does not include it\n- registry matches an older build while the collab says \"done\"\n\n### 4. End-to-End Live Probe\n\nGoal: prove the runtime actually serves post-merge behavior, not just the right metadata.\n\nUse the narrowest real probe that exercises the changed path:\n\n- actual MCP tool call\n- socket ping or request\u002Fresponse round-trip\n- CLI invocation against the real installed binary\n- HTTP endpoint hit\n- app-side behavior probe when the change is app-facing\n\nThe probe must hit the changed behavior, not just process liveness.\n\nFailure examples:\n- socket accepts connections but returns old behavior\n- MCP server starts but the changed tool path still fails\n- app bundle stamp is correct but the launched service is not the one responding\n\n## Decision Tree\n\n### 1. Merge report arrives\n\nExample inputs:\n- \"PR #264 merged\"\n- \"worker says the daemon is restarted\"\n- \"collab says deployed\"\n\nAction:\n- classify whether this is `deploy_bearing`\n- if yes, start the four-step sequence\n\n### 2. A worker says `TASK_DONE`\n\nDo not treat this as deployment proof.\n\nAction:\n- require process evidence\n- require `launchctl print`\n- require build-stamp evidence\n- require a live probe\n\nUntil those exist, the status is `UNVERIFIED DEPLOY CLAIM`.\n\n### 3. `\u002Fpost-merge-deploy-check` passes but the live probe fails\n\nVerdict:\n- `NOT DEPLOYED`\n\nReason:\n- provenance is necessary but insufficient; runtime behavior is still wrong\n\n### 4. Live probe passes but build stamp fails\n\nVerdict:\n- `DRIFTED OR NON-CANONICAL DEPLOY`\n\nReason:\n- behavior may look right now, but the agent cannot claim the merged PR is what is serving\n\n## Output Contract\n\nWhen reporting verification, use this structure:\n\n```text\nDeploy verification for \u003Crepo>#\u003Cpr>:\n- Step 1 Process replacement: PASS|FAIL — \u003Cevidence>\n- Step 2 LaunchAgent status: PASS|FAIL — \u003Cevidence>\n- Step 3 Build-stamp match: PASS|FAIL — \u003Cevidence>\n- Step 4 End-to-end probe: PASS|FAIL — \u003Cevidence>\n- Verdict: DEPLOYED | NOT DEPLOYED | NO DEPLOY REQUIRED\n```\n\nIf any evidence is missing, say `UNVERIFIED` and keep the verdict negative.\n\n## Hard Rules\n\n### 1. Merge Is Not Deploy\n\nNever answer \"done\" or \"deployed\" from GitHub merge state alone.\n\n### 2. Subagent Report Is Not Deploy\n\n`TASK_DONE`, collab notes, and \"restarted locally\" claims are not verification. `\u002Fnever-fabricate` applies in full.\n\n### 3. Provenance Alone Is Not Deploy\n\nPassing `\u002Fpost-merge-deploy-check` is not enough without the end-to-end probe.\n\n### 4. Liveness Alone Is Not Deploy\n\nA running PID or loaded LaunchAgent is not enough without build-stamp evidence.\n\n### 5. Clean Up Only After Verification\n\nDo not close the pane, remove the worktree, or report success until all required steps pass.\n\n## Anti-Patterns\n\n| Anti-pattern | Why it fails | Fix |\n|---|---|---|\n| \"PR merged, so deploy is done\" | GitHub state is not runtime state | Run all four verification steps |\n| \"worker said restarted\" | Worker self-report is not proof | Collect process, launchctl, stamp, and probe evidence |\n| \"PID exists, looks good\" | Liveness does not prove the right build or behavior | Check stamp and probe |\n| \"build stamp matches, so done\" | Provenance does not prove serving behavior | Run the live probe |\n| \"probe works, close pane\" | Probe without canonical provenance can hide drift | Run `\u002Fpost-merge-deploy-check` and record the receipt |\n\n## Receipts\n\nRecommended commands include:\n\n```bash\n: \"${ORCHESTRATOR_REPO:?ORCHESTRATOR_REPO must be set}\"\npgrep -alf '\u003Cpattern>'\nlaunchctl print gui\u002F$(id -u)\u002F\u003Cservice-label>\npython3 \"$ORCHESTRATOR_REPO\u002Fscripts\u002Fpost-merge-deploy-check.py\" \u003Crepo> \u003Cpr-number>\n\u003Ccanonical-binary> --version\n\u003Creal socket, MCP, or endpoint probe>\n```\n\nUse the canonical deploy registry when available. If the component is not registered yet, say so explicitly and still complete the process, `launchctl`, build-stamp, and live-probe checks with whatever local evidence is available.\n",{"name":5,"description":7,"disable-model-invocation":24},true,"master","3400bc971628f00ab7bf1063993477e211a5508a","skills\u002Fgolem-powers\u002Fdeploy-verify","https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems","https:\u002F\u002Fgithub.com\u002FEtanHey","etanheyman.com","2026-11-15 03:05:08",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":24},[34,37,40],{"tool":35,"command":36},"npx","npx skills add EtanHey\u002Fgolems --skill deploy-verify",{"tool":38,"command":39},"bunx","bunx skills add EtanHey\u002Fgolems --skill deploy-verify",{"tool":41,"command":42},"pnpm","pnpm dlx skills add EtanHey\u002Fgolems --skill deploy-verify","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fdeploy-verify\u002FSKILL.md",[45,51,55,59],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"evals\u002Fevals.json","REFERENCE","参考文档",7297,"https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fdeploy-verify\u002Fevals\u002Fevals.json",{"path":52,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":53,"sourceHtmlUrl":54},"evals\u002Fresults\u002Fbaseline-2026-05-02.json",977,"https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fdeploy-verify\u002Fevals\u002Fresults\u002Fbaseline-2026-05-02.json",{"path":56,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":57,"sourceHtmlUrl":58},"evals\u002Fresults\u002Fdelta-2026-05-02.json",834,"https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fdeploy-verify\u002Fevals\u002Fresults\u002Fdelta-2026-05-02.json",{"path":60,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":61,"sourceHtmlUrl":62},"evals\u002Fresults\u002Fwith-skill-2026-05-02.json",1031,"https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fdeploy-verify\u002Fevals\u002Fresults\u002Fwith-skill-2026-05-02.json",[64,71,78,85,92,99],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M7C53JEF6A6SS20Q0D2YJYM8","yash","Push dependency bugs upstream. Triggers: dependency bug, workaround, monkey-patch, fork\u002Ffix. NOT app bugs.","https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fyash\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M7C53EPMNSZKHK47VY7WQDP0","whats-new","Diff Claude\u002FCodex\u002FCursor\u002FWispr\u002Fagy changelogs vs local setup. Triggers: whats new, changelog, version bump.","https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fwhats-new\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":19},"01M7C53C4TVQ9G7AB3X57DACPB","weave","Orchestrator convergence: mine JSONLs, cite findings, route actions. Triggers: weave, run weave.","https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Fweave\u002FSKILL.md",[],[],{"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":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M7C5342EDDN7YH53KKW511V2","unslop","Cut AI tells while keeping every receipt, number, path, and hedge intact. Triggers: unslop, cut the slop, wall of text, too verbose, rambly, make this readable. NOT for delivery cadence (see i-have-adhd), channel choice, or artifact length governance.","https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Funslop\u002FSKILL.md",[],[],{"skillNo":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":19},"01M7C52M0CTDW5DDP4452FJV0G","tmp-block","Guard denying durable writes to \u002Ftmp, \u002Fprivate\u002Ftmp, \u002Fvar\u002Ffolders, $TMPDIR; enforces \u003Crepo>\u002F.worktrees\u002F\u003Cname>. Triggers: tmp guard, TMP-BLOCK deny, WORKTREE-CONVENTION deny, bypass ledger. NOT for storage policy.","https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Ftmp-block\u002FSKILL.md",[],[],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":19},"01M7C52HN4Q9E1Z74Z11K23E2H","t3-visuals","Show\u002Fvisualize\u002Fchart\u002Fdiagram\u002Fcompare\u002Fstatus board in T3 with compact interactive HTML. Not standalone dashboards.","https:\u002F\u002Fgithub.com\u002FEtanHey\u002Fgolems\u002Fblob\u002Fmaster\u002Fskills\u002Fgolem-powers\u002Ft3-visuals\u002FSKILL.md",[],[]]