[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7F2GSACKYHY9BSKM0TS9D0B":3,"skill-same:01M7F2GSACKYHY9BSKM0TS9D0B":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":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":34,"commitSha":35,"skillDirectory":36,"forksCount":37,"repositoryHtmlUrl":38,"repositoryHomepageUrl":6,"ownerHtmlUrl":39,"ownerBlogUrl":6,"ownerDescription":40,"updatedAt":41,"install":42,"resources":55},"01M7F2GSACKYHY9BSKM0TS9D0B","opencode-delegate",null,"Delegate coding tasks to the OpenCode CLI only when the user explicitly requests it, while the orchestrator retains review and landing responsibility.","lza6\u002FArmor-piercing-system-prompts","lza6","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F128880206?v=4","01M7EB3K7WPX5T28EJ2NVHE6EG","Chinese-tingfeng","01M7EB3K97E1KQ16QWKJMFNGX8",11,"MIT","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fopencode-delegate\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: opencode-delegate\ndescription: Delegate coding tasks to the OpenCode CLI only when the user explicitly\n  requests it, while the orchestrator retains review and landing responsibility.\nrisk: critical\ncategory: agent-orchestration\nsource: https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills\nsource_repo: amElnagdy\u002Fdelegate-skills\nsource_type: community\ndate_added: '2026-08-26'\nlicense: MIT\nlicense_source: https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills\u002Fblob\u002Fmaster\u002FLICENSE\ncompatibility: Requires the `opencode` CLI installed and authenticated, Node 18+,\n  and git. The orchestrating agent must be able to run shell commands and read files.\n  Shell examples assume bash\u002Fzsh (macOS\u002FLinux, or Git Bash\u002FWSL on Windows).\nmetadata:\n  version: 0.5.0\n---\n# OpenCode Delegate\n\n## When to Use\n\n- You want to delegate a bounded coding task to a separate `opencode` implementer (`OpenCode`) and then review its diff yourself.\n- The user explicitly asked for delegation to this implementer.\n\nYou are the **orchestrator**. This skill lets you hand a bounded coding task to a separate\n**implementer** — the OpenCode CLI — then review what it produced and land it yourself. You write\nthe brief and own the judgment; OpenCode does the typing in its own session; you verify and commit.\n\nNothing here is specific to one orchestrating agent. The loop needs only the ability to run a shell\ncommand and read a file, so any agent with those two capabilities — Claude Code, OpenCode driving a\nsibling session, or a comparable one — can drive it. (It is designed for and run on Claude Code; treat\nother orchestrators as designed-for, not yet proven.)\n\n## When NOT to use this\n\n- The task is small enough to just do inline — delegation overhead is not worth it.\n- The `opencode` CLI is not installed or not authenticated (run `opencode auth login`).\n- You want to write the code yourself, or you only need a review (use the `plan` agent via `--read-only`).\n\n## Prerequisites (check once)\n\n1. `opencode --version` succeeds. If not, install (`npm i -g opencode-ai`, or the native installer from\n   opencode.ai) and `opencode auth login`.\n2. **Confirm which `opencode` is on PATH.** `command -v opencode` shows the active binary and\n   `opencode --version` its version. The relay records the version it ran into `result.json`, so a stale\n   binary is visible after the fact.\n3. A model provider is authenticated — `opencode auth list` shows at least one credential.\n4. You are in (or will point `--cd` at) the target git repository.\n\n## Choose the implementer model\n\nOpenCode has **no safe default** — a bare `opencode run` errors — so a fresh run needs a model via\n`--model` or a fleet `--lane` that sets one (a resumed run inherits its session's model). Naming the\nmodel is the one decision a single-model backend like codex-delegate never had, and it has two owners:\n\n- **The human owns which models are allowed.** `opencode models` lists hundreds of entries, most billed\n  per token (OpenRouter and the like); only the human knows which are their flat-rate subscriptions, and\n  the CLI can't tell them apart. So the usable set is theirs — ideally stated once in the repo's\n  `AGENTS.md` or their `CLAUDE.md` (e.g. \"delegate mechanical work to `opencode-go\u002F…`, hard logic to\n  `…`\").\n- **You, the orchestrator, pick per task — from that set.** Match the model to the brief: a cheap, fast\n  model for a mechanical sweep (rename, migration, removal); a strong one for a subtle bug or a\n  money\u002Fsecurity path.\n- **If no usable set is stated, ask — don't guess.** Guessing from the catalog risks a metered model and\n  a surprise bill. Name the constraint to the human and let them choose.\n\nMore depth: [references\u002Fwriting-the-brief.md](references\u002Fwriting-the-brief.md).\n\n## The loop\n\nRun these five steps per task. Steps 1, 4, and 5 are your judgment; 2 and 3 are mechanical.\n\n### 1. Write the brief\n\nOpenCode sees **only** the text you send plus what it can read from the working tree — no chat history,\nno shared context. Everything the task needs goes in the brief: the goal, the current state, what to\nchange, what to leave untouched, the project's **actual** gate commands (discover them from the repo's\nAGENTS.md\u002FCLAUDE.md\u002FMakefile — do not assume), and a report contract. Tell OpenCode it will **not**\ncommit (you will). Keep one task per brief. Full guidance and a template:\n[references\u002Fwriting-the-brief.md](references\u002Fwriting-the-brief.md).\n\n### 2. Dispatch\n\nSend the brief to OpenCode with the bundled helper. It wraps `opencode run`, captures the run, and\nwrites a structured `result.json` — so your only job is \"run a command, read a file.\" (`\u003Cskill-dir>`\nbelow is this skill's installed directory — the folder containing this `SKILL.md`. Claude Code prints\nit as \"Base directory for this skill\" when the skill loads; on other orchestrators use that same\ndirectory — if unsure where it landed, run `find ~ -name relay.mjs -path '*opencode-delegate*'` and\nsubstitute the directory above it.)\n\n```bash\nnode \"\u003Cskill-dir>\u002Fscripts\u002Frelay.mjs\" --brief brief.txt --model \u003Cprovider\u002Fmodel> --cd \u002Fpath\u002Fto\u002Frepo\n# --model (or a --lane that sets model) is required on a fresh run\n# fleet lane from delegate-setup:           add --lane \u003Cname>  (dials apply; flags still win)\n# read-only (review\u002Fdiagnosis, no edits):   add --read-only   (uses the plan agent)\n# continue the previous OpenCode session:   add --resume-last  (delta brief only; keeps the model)\n# hard time limit (watchdog):               add --timeout 2h  (default: off; implementation runs routinely need 1-2h)\n# see all options:                          node ...\u002Frelay.mjs --help\n```\n\nThe helper defaults to the write-capable `build` agent and writes its artifacts to a temp dir, so the\nrepo under review stays clean. It **never commits** — see step 5. Mechanics, flags, and the\n`result.json` shape: [references\u002Fdispatch-and-poll.md](references\u002Fdispatch-and-poll.md).\n\n### 3. Wait for completion\n\nThe helper blocks until OpenCode finishes, so back it with whatever your orchestrator offers and resume\nwhen it returns:\n\n- **Claude Code:** run the Bash call with `run_in_background: true`; you are notified on completion.\n- **Plain shell \u002F other agents:** run it in the foreground for short tasks, or background it and poll\n  the result file — `… &` in bash\u002Fzsh (including Git Bash\u002FWSL), or your shell's equivalent (`Start-Job`\n  in PowerShell, `start \u002Fb` in cmd). The run is done when `result.json` exists with a `status`. (A\n  pre-run usage error — bad args or an empty brief — instead exits with code 2 and writes no result\n  file, so check the exit code too. A missing `opencode` binary exits 127 but *does* write a\n  `result.json` with status `opencode_unavailable`.)\n\nDo not trust progress trackers over reality: a run is finished when `result.json` is written and the\nprocess has exited. Read the working tree, not a status line. The implementer's full report is\nthe `finalMessage` field in `result.json` (also printed in full on stdout between the report markers).\n\n### 4. Review — do not trust the self-report\n\nOpenCode's `result.json` includes its own final message and any gate claims. **Re-verify, don't accept:**\n\n- **Re-run the project's gates yourself** (the test\u002Flint\u002Fbuild commands from step 1). Never take\n  \"gates passed\" on faith.\n- **Read the diff** against the brief: did OpenCode do what was asked, nothing more (scope creep) and\n  nothing less? `touchedFiles` in the result is your starting point.\n- **Run the relevant guard skills** on the diff if you have them installed (clean-code-guard,\n  test-guard, etc. from `guard-skills`) — this skill produces the work; those skills judge it.\n- For schema\u002Fmigration changes, round-trip them; for removals, grep for dangling references.\n\nFull checklist: [references\u002Freview-and-land.md](references\u002Freview-and-land.md).\n\n### 5. Land it\n\nThe implementer edits the working tree; **the orchestrator commits.** Committing should be the act of\nthe party that verified the work. Only after the gates pass and the diff holds:\n\n- Commit the verified work yourself, with a clear message.\n- If it needs changes, send a delta brief with `--resume-last` (don't restate the whole task) and\n  review again.\n\n## Autonomy model\n\nOpenCode's autonomy is governed by the **agent**, not a sandbox enum:\n\n- **`build`** (the relay default) — write-capable; edits files in the working dir headlessly. The\n  equivalent of \"let it implement.\"\n- **`plan`** (via `--read-only`) — read-only; reviews and diagnoses without touching the tree. The\n  equivalent of \"let it look but not edit.\"\n\nPermissions **auto-approve by default**: the relay passes `--auto` so a headless run never blocks on a\nprompt no one can answer. That is the point of unattended delegation — the orchestrator's diff review\nand the implementer sweep (step 4) are the safety net, not a per-action prompt. Pass `--no-auto` to\nhonor the agent's own permission config instead (allow\u002Fask\u002Fdeny per action); pair it with an agent whose\nin-workspace permissions are set to *allow*, or a headless run can hang waiting on an `ask`.\n**Read-only (`plan`) runs never get `--auto`** — auto-approving would let the plan agent's ask-gated\nedit\u002Fbash permissions through and defeat \"read-only,\" so a review can't be tricked into touching the tree.\n\n## Authorization model\n\nDelegation is something the human opts into. Once they have (\"run this queue\", \"proceed\"), committing\nverified, gate-passing work is the agreed contract — that is the whole point. Two limits on that\nmandate: **surface, don't absorb** (report OpenCode's design decisions, defensible-but-unasked turns,\nand non-blocking nitpicks rather than silently keeping them) and **stop for scope changes** (if correct\ncompletion needs going beyond the brief, ask — don't expand the mandate yourself). The full treatment\nis in [references\u002Freview-and-land.md](references\u002Freview-and-land.md).\n\n## References\n\n- [references\u002Fwriting-the-brief.md](references\u002Fwriting-the-brief.md) — how to write a brief OpenCode can\n  execute blind: structure, XML blocks, the report contract, embedding the real gate commands.\n- [references\u002Fdispatch-and-poll.md](references\u002Fdispatch-and-poll.md) — `relay.mjs` flags, the\n  `result.json` contract, backgrounding per orchestrator, and recovery when a run misbehaves.\n- [references\u002Freview-and-land.md](references\u002Freview-and-land.md) — the review checklist, the commit\n  boundary, and the rework cycle via `--resume-last`.\n- [references\u002Fmulti-task-queues.md](references\u002Fmulti-task-queues.md) — running a sequential queue:\n  carrying constraints forward, progress tracking, and the end-of-run coherence check.\n\n\n## Limitations\n\n- Docs-only import — executable `scripts\u002Frelay.mjs` not included; see upstream for full runtime. Requires `opencode` CLI, Node 18+, git.\n- Relay never commits — it only returns structured result JSON; you review and land the commit.\n\n> Adapted from [amElnagdy\u002Fdelegate-skills](https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills) (MIT) — docs-only, runtime not bundled.\n","# OpenCode Delegate\n\n## When to Use\n\n- You want to delegate a bounded coding task to a separate `opencode` implementer (`OpenCode`) and then review its diff yourself.\n- The user explicitly asked for delegation to this implementer.\n\nYou are the **orchestrator**. This skill lets you hand a bounded coding task to a separate\n**implementer** — the OpenCode CLI — then review what it produced and land it yourself. You write\nthe brief and own the judgment; OpenCode does the typing in its own session; you verify and commit.\n\nNothing here is specific to one orchestrating agent. The loop needs only the ability to run a shell\ncommand and read a file, so any agent with those two capabilities — Claude Code, OpenCode driving a\nsibling session, or a comparable one — can drive it. (It is designed for and run on Claude Code; treat\nother orchestrators as designed-for, not yet proven.)\n\n## When NOT to use this\n\n- The task is small enough to just do inline — delegation overhead is not worth it.\n- The `opencode` CLI is not installed or not authenticated (run `opencode auth login`).\n- You want to write the code yourself, or you only need a review (use the `plan` agent via `--read-only`).\n\n## Prerequisites (check once)\n\n1. `opencode --version` succeeds. If not, install (`npm i -g opencode-ai`, or the native installer from\n   opencode.ai) and `opencode auth login`.\n2. **Confirm which `opencode` is on PATH.** `command -v opencode` shows the active binary and\n   `opencode --version` its version. The relay records the version it ran into `result.json`, so a stale\n   binary is visible after the fact.\n3. A model provider is authenticated — `opencode auth list` shows at least one credential.\n4. You are in (or will point `--cd` at) the target git repository.\n\n## Choose the implementer model\n\nOpenCode has **no safe default** — a bare `opencode run` errors — so a fresh run needs a model via\n`--model` or a fleet `--lane` that sets one (a resumed run inherits its session's model). Naming the\nmodel is the one decision a single-model backend like codex-delegate never had, and it has two owners:\n\n- **The human owns which models are allowed.** `opencode models` lists hundreds of entries, most billed\n  per token (OpenRouter and the like); only the human knows which are their flat-rate subscriptions, and\n  the CLI can't tell them apart. So the usable set is theirs — ideally stated once in the repo's\n  `AGENTS.md` or their `CLAUDE.md` (e.g. \"delegate mechanical work to `opencode-go\u002F…`, hard logic to\n  `…`\").\n- **You, the orchestrator, pick per task — from that set.** Match the model to the brief: a cheap, fast\n  model for a mechanical sweep (rename, migration, removal); a strong one for a subtle bug or a\n  money\u002Fsecurity path.\n- **If no usable set is stated, ask — don't guess.** Guessing from the catalog risks a metered model and\n  a surprise bill. Name the constraint to the human and let them choose.\n\nMore depth: [references\u002Fwriting-the-brief.md](references\u002Fwriting-the-brief.md).\n\n## The loop\n\nRun these five steps per task. Steps 1, 4, and 5 are your judgment; 2 and 3 are mechanical.\n\n### 1. Write the brief\n\nOpenCode sees **only** the text you send plus what it can read from the working tree — no chat history,\nno shared context. Everything the task needs goes in the brief: the goal, the current state, what to\nchange, what to leave untouched, the project's **actual** gate commands (discover them from the repo's\nAGENTS.md\u002FCLAUDE.md\u002FMakefile — do not assume), and a report contract. Tell OpenCode it will **not**\ncommit (you will). Keep one task per brief. Full guidance and a template:\n[references\u002Fwriting-the-brief.md](references\u002Fwriting-the-brief.md).\n\n### 2. Dispatch\n\nSend the brief to OpenCode with the bundled helper. It wraps `opencode run`, captures the run, and\nwrites a structured `result.json` — so your only job is \"run a command, read a file.\" (`\u003Cskill-dir>`\nbelow is this skill's installed directory — the folder containing this `SKILL.md`. Claude Code prints\nit as \"Base directory for this skill\" when the skill loads; on other orchestrators use that same\ndirectory — if unsure where it landed, run `find ~ -name relay.mjs -path '*opencode-delegate*'` and\nsubstitute the directory above it.)\n\n```bash\nnode \"\u003Cskill-dir>\u002Fscripts\u002Frelay.mjs\" --brief brief.txt --model \u003Cprovider\u002Fmodel> --cd \u002Fpath\u002Fto\u002Frepo\n# --model (or a --lane that sets model) is required on a fresh run\n# fleet lane from delegate-setup:           add --lane \u003Cname>  (dials apply; flags still win)\n# read-only (review\u002Fdiagnosis, no edits):   add --read-only   (uses the plan agent)\n# continue the previous OpenCode session:   add --resume-last  (delta brief only; keeps the model)\n# hard time limit (watchdog):               add --timeout 2h  (default: off; implementation runs routinely need 1-2h)\n# see all options:                          node ...\u002Frelay.mjs --help\n```\n\nThe helper defaults to the write-capable `build` agent and writes its artifacts to a temp dir, so the\nrepo under review stays clean. It **never commits** — see step 5. Mechanics, flags, and the\n`result.json` shape: [references\u002Fdispatch-and-poll.md](references\u002Fdispatch-and-poll.md).\n\n### 3. Wait for completion\n\nThe helper blocks until OpenCode finishes, so back it with whatever your orchestrator offers and resume\nwhen it returns:\n\n- **Claude Code:** run the Bash call with `run_in_background: true`; you are notified on completion.\n- **Plain shell \u002F other agents:** run it in the foreground for short tasks, or background it and poll\n  the result file — `… &` in bash\u002Fzsh (including Git Bash\u002FWSL), or your shell's equivalent (`Start-Job`\n  in PowerShell, `start \u002Fb` in cmd). The run is done when `result.json` exists with a `status`. (A\n  pre-run usage error — bad args or an empty brief — instead exits with code 2 and writes no result\n  file, so check the exit code too. A missing `opencode` binary exits 127 but *does* write a\n  `result.json` with status `opencode_unavailable`.)\n\nDo not trust progress trackers over reality: a run is finished when `result.json` is written and the\nprocess has exited. Read the working tree, not a status line. The implementer's full report is\nthe `finalMessage` field in `result.json` (also printed in full on stdout between the report markers).\n\n### 4. Review — do not trust the self-report\n\nOpenCode's `result.json` includes its own final message and any gate claims. **Re-verify, don't accept:**\n\n- **Re-run the project's gates yourself** (the test\u002Flint\u002Fbuild commands from step 1). Never take\n  \"gates passed\" on faith.\n- **Read the diff** against the brief: did OpenCode do what was asked, nothing more (scope creep) and\n  nothing less? `touchedFiles` in the result is your starting point.\n- **Run the relevant guard skills** on the diff if you have them installed (clean-code-guard,\n  test-guard, etc. from `guard-skills`) — this skill produces the work; those skills judge it.\n- For schema\u002Fmigration changes, round-trip them; for removals, grep for dangling references.\n\nFull checklist: [references\u002Freview-and-land.md](references\u002Freview-and-land.md).\n\n### 5. Land it\n\nThe implementer edits the working tree; **the orchestrator commits.** Committing should be the act of\nthe party that verified the work. Only after the gates pass and the diff holds:\n\n- Commit the verified work yourself, with a clear message.\n- If it needs changes, send a delta brief with `--resume-last` (don't restate the whole task) and\n  review again.\n\n## Autonomy model\n\nOpenCode's autonomy is governed by the **agent**, not a sandbox enum:\n\n- **`build`** (the relay default) — write-capable; edits files in the working dir headlessly. The\n  equivalent of \"let it implement.\"\n- **`plan`** (via `--read-only`) — read-only; reviews and diagnoses without touching the tree. The\n  equivalent of \"let it look but not edit.\"\n\nPermissions **auto-approve by default**: the relay passes `--auto` so a headless run never blocks on a\nprompt no one can answer. That is the point of unattended delegation — the orchestrator's diff review\nand the implementer sweep (step 4) are the safety net, not a per-action prompt. Pass `--no-auto` to\nhonor the agent's own permission config instead (allow\u002Fask\u002Fdeny per action); pair it with an agent whose\nin-workspace permissions are set to *allow*, or a headless run can hang waiting on an `ask`.\n**Read-only (`plan`) runs never get `--auto`** — auto-approving would let the plan agent's ask-gated\nedit\u002Fbash permissions through and defeat \"read-only,\" so a review can't be tricked into touching the tree.\n\n## Authorization model\n\nDelegation is something the human opts into. Once they have (\"run this queue\", \"proceed\"), committing\nverified, gate-passing work is the agreed contract — that is the whole point. Two limits on that\nmandate: **surface, don't absorb** (report OpenCode's design decisions, defensible-but-unasked turns,\nand non-blocking nitpicks rather than silently keeping them) and **stop for scope changes** (if correct\ncompletion needs going beyond the brief, ask — don't expand the mandate yourself). The full treatment\nis in [references\u002Freview-and-land.md](references\u002Freview-and-land.md).\n\n## References\n\n- [references\u002Fwriting-the-brief.md](references\u002Fwriting-the-brief.md) — how to write a brief OpenCode can\n  execute blind: structure, XML blocks, the report contract, embedding the real gate commands.\n- [references\u002Fdispatch-and-poll.md](references\u002Fdispatch-and-poll.md) — `relay.mjs` flags, the\n  `result.json` contract, backgrounding per orchestrator, and recovery when a run misbehaves.\n- [references\u002Freview-and-land.md](references\u002Freview-and-land.md) — the review checklist, the commit\n  boundary, and the rework cycle via `--resume-last`.\n- [references\u002Fmulti-task-queues.md](references\u002Fmulti-task-queues.md) — running a sequential queue:\n  carrying constraints forward, progress tracking, and the end-of-run coherence check.\n\n\n## Limitations\n\n- Docs-only import — executable `scripts\u002Frelay.mjs` not included; see upstream for full runtime. Requires `opencode` CLI, Node 18+, git.\n- Relay never commits — it only returns structured result JSON; you review and land the commit.\n\n> Adapted from [amElnagdy\u002Fdelegate-skills](https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills) (MIT) — docs-only, runtime not bundled.\n",{"name":5,"risk":24,"source":25,"license":15,"category":26,"metadata":27,"date_added":29,"description":7,"source_repo":30,"source_type":31,"compatibility":32,"license_source":33},"critical","https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills","agent-orchestration",{"version":28},"0.5.0","2026-08-26","amElnagdy\u002Fdelegate-skills","community","Requires the `opencode` CLI installed and authenticated, Node 18+, and git. The orchestrating agent must be able to run shell commands and read files. Shell examples assume bash\u002Fzsh (macOS\u002FLinux, or Git Bash\u002FWSL on Windows).","https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills\u002Fblob\u002Fmaster\u002FLICENSE","main","d69096e239f2a89aabc2cd078c47664a9d3b1e11","05-技能与代理开发\u002Fopencode-delegate",3,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts","https:\u002F\u002Fgithub.com\u002Flza6","个人vx：Tf00798\r\n个人✈：@TFGY999","2026-10-08 14:29:22",{"source":8,"skillName":5,"commands":43,"prompt":53,"zipAvailable":54},[44,47,50],{"tool":45,"command":46},"npx","npx skills add lza6\u002FArmor-piercing-system-prompts --skill opencode-delegate",{"tool":48,"command":49},"bunx","bunx skills add lza6\u002FArmor-piercing-system-prompts --skill opencode-delegate",{"tool":51,"command":52},"pnpm","pnpm dlx skills add lza6\u002FArmor-piercing-system-prompts --skill opencode-delegate","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fopencode-delegate\u002FSKILL.md",true,[56,62,66,70],{"path":57,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Fdispatch-and-poll.md","REFERENCE","参考文档",10161,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fopencode-delegate\u002Freferences\u002Fdispatch-and-poll.md",{"path":63,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Fmulti-task-queues.md",3928,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fopencode-delegate\u002Freferences\u002Fmulti-task-queues.md",{"path":67,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":68,"sourceHtmlUrl":69},"references\u002Freview-and-land.md",7851,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fopencode-delegate\u002Freferences\u002Freview-and-land.md",{"path":71,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":72,"sourceHtmlUrl":73},"references\u002Fwriting-the-brief.md",7524,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fopencode-delegate\u002Freferences\u002Fwriting-the-brief.md",[75,82,87,94,101,108],{"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":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7FFG485KY83TD6HSC4YPCMC","obsidian-clipper-template-creator","Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator\u002FSKILL.md",[],[],{"skillNo":83,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":19},"01M7FFG0T6AZ0STEJPSB3CX84K","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator-antigravity-awesome-skills-main\u002FSKILL.md",[],[],{"skillNo":88,"name":89,"nameZh":6,"description":90,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":19},"01M7FFFY0QEM14SKWPDYGMXM2P","python-api","Programmatic Object-Detection-Metrics API guidance for BoundingBox, BoundingBoxes, Evaluator, enums, coordinate conversion, AP\u002FmAP semantics, and source-style import diagnostics.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002Fsub-skills\u002Fpython-api\u002FSKILL.md",[],[],{"skillNo":95,"name":96,"nameZh":6,"description":97,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":19},"01M7FFFVZQZ2GJ0C0TGP7JHD20","file-evaluation","Folder-based PASCAL VOC AP\u002FmAP evaluation from ground-truth and detection text files.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002Fsub-skills\u002Ffile-evaluation\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":15,"licenseFileUrl":6,"sourceHtmlUrl":105,"categories":106,"keywords":107,"publishedAt":19},"01M7FFFSFNYSE77SJ6PE1CK2DA","object-detection-metrics","Guides legacy Object-Detection-Metrics PASCAL VOC AP\u002FmAP evaluation from detection text folders and direct Python API classes.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002FSKILL.md",[],[],{"skillNo":109,"name":110,"nameZh":6,"description":111,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":112,"categories":113,"keywords":114,"publishedAt":19},"01M7FFFP0GCK1RVXBCPQ8APBMT","nx-workspace-patterns","Configure and optimize Nx monorepo workspaces. Use when setting up Nx, configuring project boundaries, optimizing build caching, or implementing affected commands.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fnx-workspace-patterns\u002FSKILL.md",[],[]]