[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CWPC7NHJXK7XW7Z65HBNFX":3,"skill-same:01M7CWPC7NHJXK7XW7Z65HBNFX":48},{"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":26,"commitSha":27,"skillDirectory":28,"forksCount":14,"repositoryHtmlUrl":29,"repositoryHomepageUrl":6,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7CWPC7NHJXK7XW7Z65HBNFX","loop-design-check",null,"Design a goal-oriented agent loop, and review it for the ways loops go wrong — spinning and burning tokens, Goodhart-gaming the verifier, or running a wrong answer to completion. Two actions: (1) WRITE a loop — gate whether to build it, define a machine-decidable goal, pick the loop type, pick a skeleton; (2) REVIEW a loop — run it past five failure modes plus decidability, boundaries, fallback, judge independence, and keep-judgment-with-the-human red lines. Use when designing an autonomous agent loop, or when you already have one and worry it will spin, cheat, or run a wrong answer to the end. Complements the mechanism-layer loop skills (autonomous-loops, continuous-agent-loop) by covering the judgment layer they don't. 中文触发：写 loop、设计 loop、做一个 loop、检查 loop 对不对、loop 体检、loop 会不会跑飞、可判定目标、五个崩法、plan build judge。English triggers: design an agent loop, write a loop, check a loop, loop review, prevent a runaway loop, goal-oriented loop, decidable goal, plan\u002Fbuild\u002Fjudge.","Henrique1601\u002FSkiilsToIAs","Henrique1601","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F92006691?v=4","01M7CRXVT0GVXVK80JT1GR687S","Henrique Bezerra Dos Santos","01M7CRXVTN5Y8TMWJ1MDSKCQ1T",0,"MIT","https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs\u002Fblob\u002Fmain\u002Futilities\u002Floop-design-check\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: loop-design-check\ndescription: \"Design a goal-oriented agent loop, and review it for the ways loops go wrong — spinning and burning tokens, Goodhart-gaming the verifier, or running a wrong answer to completion. Two actions: (1) WRITE a loop — gate whether to build it, define a machine-decidable goal, pick the loop type, pick a skeleton; (2) REVIEW a loop — run it past five failure modes plus decidability, boundaries, fallback, judge independence, and keep-judgment-with-the-human red lines. Use when designing an autonomous agent loop, or when you already have one and worry it will spin, cheat, or run a wrong answer to the end. Complements the mechanism-layer loop skills (autonomous-loops, continuous-agent-loop) by covering the judgment layer they don't. 中文触发：写 loop、设计 loop、做一个 loop、检查 loop 对不对、loop 体检、loop 会不会跑飞、可判定目标、五个崩法、plan build judge。English triggers: design an agent loop, write a loop, check a loop, loop review, prevent a runaway loop, goal-oriented loop, decidable goal, plan\u002Fbuild\u002Fjudge.\"\nmetadata:\n  origin: ECC\n---\n\n# Loop Design + Review\n\n> **Premise.** An LLM is a feed-forward system: prompt in → tokens out, with no built-in \"steer toward the goal\" across turns. To make it *behave* like a goal-oriented system, you wrap a feedback loop around it. This skill helps you **write** that loop correctly and **review** it so it won't run away.\n\n## When to use \u002F not\n\n**Use it when:**\n- You want to hand a repeating task to an agent that runs over and over (write→test, test→fix, fix→verify…).\n- You already have a loop and worry it spins, cheats, or runs a wrong answer to completion.\n\n**Don't use it for:**\n- A one-off task → just do it; don't wrap a loop around it.\n- A plain timer \u002F poll → use `\u002Floop`; no design needed.\n- *How to wire the loop architecture* (pipelines → DAGs, long-run recovery) → that's the mechanism layer; see `autonomous-loops` \u002F `continuous-agent-loop`. **This skill only covers \"is the goal right, and will it run away\" — it does not re-explain mechanism.**\n\n## Red-line premise: two levels of feedback\n\n| Level | Who owns it | What it does |\n|---|---|---|\n| **Execution** (low) | machine \u002F agent | Measures \"how far from the literal goal\" and grinds it to zero. The machine is strong here. |\n| **Judgment** (high) | **human** | Decides \"is this goal itself right, should it change, should it stop.\" The machine can't step outside its own loop to question the goal. |\n\n> A thermostat can feed back \"how far from 26°C,\" but when you have a fever and want 28°C it can't judge whether 26 is the *right* target — it just grinds toward 26. **\"What to set today\" is always the human's call.**\n> Handing judgment \u002F sign-off \u002F the last switch to the machine = removing the high-level feedback = it sprints, fast and hard, toward a goal no one questioned → wrong output.\n\n---\n\n## Action 1 — Write a loop (5 steps)\n\n### Step 0 · Subtract first: should you even build it? (4-condition gate, any miss = veto)\n\n① the task repeats weekly or more　② verification can be automated　③ the token budget can take it　④ the agent has tools that actually *run and see the result*\n\nMiss any one → **don't build a loop**; do it by hand or another way.\n> What stops most people isn't \"can I write a loop,\" it's \"does my repo deserve one.\" A repo that deserves a loop has a reconciliation baseline (golden sample \u002F upstream total) + tests + a lint guard. **A repo that doesn't deserve a loop will only have its errors amplified by one.**\n\n### Step 1 · Define a *machine-decidable* goal (the hard part — the loop lives or dies here)\n\nThe whole loop rides on the comparator's \"is it done yet?\" **The comparator can only work if your exit condition can be judged yes\u002Fno by a machine.**\n\n- Bad: Vague (\"make it good,\" \"write it sharper\") → the comparator can't judge → either it never passes (stuck retrying) or it guesses (passes\u002Fblocks at random).\n- Good: Decidable (\"all 96 unit tests green AND a change-list is produced,\" \"module-02 fields filled, pytest passes, business logic untouched\") → one check settles it; the loop converges cleanly.\n\n**Five-point goal framework:**\n1. **Done-criterion is machine-verifiable.**\n2. **Boundary conditions defined alongside the done-criterion** (\"what it must NOT do\") — anti-Goodhart; missing boundaries = a license to cheat.\n3. **Has a failure fallback** — retry cap N + escalate to a human when exceeded.\n4. **Goal is layered.**\n5. **Prefer reconciliation over assertion for the done-criterion** — anchor to external fact (golden sample \u002F upstream total \u002F financial tie-out \u002F platform back-office numbers) before your own assertions. \"All tests pass\" can be gamed (loosen asserts, fake mocks, swallow exceptions); \"diff vs the reference \u003C 0.01\" can't.\n\n> **Self-check:** read the goal to someone who doesn't know the domain — can they run one command and tell whether it's done? If not, it isn't decidable enough. Go back.\n\n### Step 2 · Pick the loop type\n\n| Your task | Loop type (cybernetic) | How it stops |\n|---|---|---|\n| Has a clear \"done\" test (write to done \u002F a batch of images processed) | **servo** (`\u002Fgoal`-style closed-loop) | stops on reaching the goal |\n| No endpoint, must keep maintaining a state (inventory alert \u002F scheduled health check) | **regulator** (`\u002Floop`-style thermostat) | never stops; acts only on change (dead-band suppresses noise) |\n| Periodic sampling, stop on a condition (watch a PR until CI is green) | **regulator with an exit** | stops when the exit condition holds |\n| Must \"ensure something happens on time\" | wrap the above in `\u002Fschedule` | cron fires it |\n\n> Rule of thumb: clear \"done\" test → servo; must keep maintaining, no endpoint → regulator; must \"happen on time\" → wrap a regulator in schedule.\n\n### Step 3 · Pick a skeleton\n\n**Maintenance type (tend something that exists) → document-driven dispatch.**\nThe loop isn't \"run a fixed check on a timer,\" it's **\"read a doc on a timer, and dispatch only when the doc changed.\"** The doc is the task queue + state machine + human interface.\nThree disciplines: ① the problem column is human-write-only, the result column is loop-write-only, **state advances one-way and never rolls back**; ② **the exit code is final** (if the script says exit 1, the script wins); ③ state advances only as far as \"awaiting verification\" — **the \"done\" cell is flipped by a human only.** The loop is the worker, not the acceptance officer.\n\n**Greenfield type (build from scratch) → plan \u002F build \u002F judge, three roles.**\n\n| Role | Does | Key |\n|---|---|---|\n| **Plan** | break the goal into a spec + **decidable acceptance conditions** | acceptance must be script-judgeable |\n| **Build** | write to the spec | **must not change the acceptance conditions** |\n| **Judge** | run acceptance **independently**; pass → stop, fail → return with the failure reason to Build | **independent + deterministic** |\n\nThree iron rules (all bet on the judge): ① **the judge must be independent** — not the same agent as Build (grading your own homework always inflates); ② **deterministic rules** — pytest \u002F reconciliation diff \u002F type check \u002F diff, never \"looks right\"; ③ **Build may not edit the acceptance conditions to pass**. Three failed retries → escalate to a human.\n\n### Step 4 · Add damping (against oscillation \u002F runaway)\n\nRetry cap, hard stop, human flips the last switch = damping. **Negative feedback with no damping oscillates** (the Ralph-Wiggum loop: spinning in place, burning tokens).\n\n### Step 5 · Land in three stages (don't go fully automatic on day one)\n\n① **Run it once by hand** (forces you to state exactly \"how the judge decides\") → ② harden into a skill \u002F Claude Code sub-agents (a main Claude loops, dispatching plan\u002Fbuild\u002Fjudge) → ③ hang it on cron for full automation.\n\n---\n\n## Action 2 — Review a loop (checklist = five failure modes)\n\n> Run the loop past each row. **Hitting any one = this loop will misfire; send it back.** These five are negative experience (gotchas) — worth more than positive rules.\n\n| # | Failure mode (how it breaks) | Review question (a hit = red) | Antibody |\n|---|---|---|---|\n| 1 | Goal is a correct platitude → **spins, burns money** | Can the exit condition be machine-judged yes\u002Fno? Or is it \"manage it well \u002F make it good\"? | Replace with a decidable result condition (Action 1·Step 1) |\n| 2 | \"Verification\" written as \"check if it looks ok\" → **agent confidently says fine and stops** | Is the judge the defendant itself? Does verification rest on \"looks right\" or deterministic rules? | Reconcile + exit code rules + independent judge |\n| 3 | (worst) Only gates on \"all tests pass\" → **agent deletes the tests** | Is there a boundary (\"what it must NOT do\")? Or only a done-criterion? | Done-criterion **+ boundary** together (the Goodhart antibody) |\n| 4 | Counts on the agent asking mid-run → **it won't; it runs the wrong answer to the end** | Is there any \"clarify only at runtime\" point? | **Front-load every clarification**; settle it once before launch |\n| 5 | Bloated CLAUDE.md + stale memory → **the faster it loops, the more it errs** | Are the docs\u002Fmemory it depends on fresh? Who maintains them? | Layered memory + periodic lint |\n\n**Plus three red lines (violate any = not allowed to go automatic):**\n- **Keep judgment with the human.** Acceptance \u002F the \"done\" cell is flipped by a human; the loop is not the acceptance officer.\n- **Responsibility doesn't transfer.** Anything whose failure you can't afford (merge the wrong PR \u002F publish the wrong thing \u002F misallocate money) → **don't hand over the authority automatically.**\n- **Counter-intuitive warning.** The more \"self-improving \u002F rewrites-its-own-rules\" a loop is, the **stricter the human review it needs** (to see what it rewrote the rules into) — not looser. The machine is too fast to intercept after the fact, so the human's judgment must sit **before the action** (a hard gate), not as a post-hoc patch.\n\n---\n\n## Worked example — reviewing a \"nightly green-keeper\" loop\n\nYou want a loop that runs every night and fixes whatever tests are failing.\n\n- **Naive goal:** \"make all tests pass.\" → Step-1 self-check fails: this is the bait for failure mode #3.\n- **Decidable goal (fixed):** \"all tests green **AND** no test file deleted or weakened **AND** coverage not lowered **AND** a change-list produced.\" Boundary now defined alongside the done-criterion.\n- **Type:** servo with a retry cap of 3 (Step 2 + Step 4).\n- **Skeleton:** plan\u002Fbuild\u002Fjudge — the **judge is CI run independently**, never the fixing agent (Step 3).\n\nNow run the **review checklist**, and it catches what the naive version would have missed:\n- **#3 hit** → the naive \"all tests pass\" lets the agent delete a failing test to \"win.\" Fixed by the boundary \"no test file deleted\u002Fweakened.\"\n- **#2 hit** → if the fixing agent also judged its own fix, it would pass itself. Fixed by \"judge = independent CI, deterministic.\"\n- **#4 hit** → if a fix is ambiguous, the agent won't stop to ask at 2 a.m.; it'll commit a guess. Fixed by front-loading: ambiguous fixes are left for the human, not guessed.\n- **Red line** → the loop opens a PR but **does not auto-merge**; the human flips the last switch (responsibility doesn't transfer).\n\nThe naive loop and the reviewed loop differ by four lines of constraint — and that's the difference between \"wakes you to a deleted test suite\" and \"wakes you to a clean PR.\"\n\n---\n\n## One-line close\n\n> The hard part of writing a loop isn't \"can I write a loop,\" it's **defining a goal a machine can reconcile** — decidable, bounded, reconciliation-based. The controller must be deterministic and external; keep judgment and the standard with the human; the system tends toward entropy, so maintain it.\n> **A loop only rewards someone who has already thought it through. Count on it to think for you, and it will happily think wrong, with you, at scale.**\n\n---\n\n> Lineage: Wiener's two-level feedback (*The Human Use of Human Beings*, 1950) for the judgment\u002Fexecution split and red lines; the plan\u002Fbuild\u002Fjudge pattern from Anatoli's *Loops explained* and Addy's *Loop Engineering*.\n> Mechanism layer (how to wire the loop architecture): see `autonomous-loops` \u002F `continuous-agent-loop`. This skill does not re-implement mechanism; it covers goal definition and runaway prevention only.\n","# Loop Design + Review\n\n> **Premise.** An LLM is a feed-forward system: prompt in → tokens out, with no built-in \"steer toward the goal\" across turns. To make it *behave* like a goal-oriented system, you wrap a feedback loop around it. This skill helps you **write** that loop correctly and **review** it so it won't run away.\n\n## When to use \u002F not\n\n**Use it when:**\n- You want to hand a repeating task to an agent that runs over and over (write→test, test→fix, fix→verify…).\n- You already have a loop and worry it spins, cheats, or runs a wrong answer to completion.\n\n**Don't use it for:**\n- A one-off task → just do it; don't wrap a loop around it.\n- A plain timer \u002F poll → use `\u002Floop`; no design needed.\n- *How to wire the loop architecture* (pipelines → DAGs, long-run recovery) → that's the mechanism layer; see `autonomous-loops` \u002F `continuous-agent-loop`. **This skill only covers \"is the goal right, and will it run away\" — it does not re-explain mechanism.**\n\n## Red-line premise: two levels of feedback\n\n| Level | Who owns it | What it does |\n|---|---|---|\n| **Execution** (low) | machine \u002F agent | Measures \"how far from the literal goal\" and grinds it to zero. The machine is strong here. |\n| **Judgment** (high) | **human** | Decides \"is this goal itself right, should it change, should it stop.\" The machine can't step outside its own loop to question the goal. |\n\n> A thermostat can feed back \"how far from 26°C,\" but when you have a fever and want 28°C it can't judge whether 26 is the *right* target — it just grinds toward 26. **\"What to set today\" is always the human's call.**\n> Handing judgment \u002F sign-off \u002F the last switch to the machine = removing the high-level feedback = it sprints, fast and hard, toward a goal no one questioned → wrong output.\n\n---\n\n## Action 1 — Write a loop (5 steps)\n\n### Step 0 · Subtract first: should you even build it? (4-condition gate, any miss = veto)\n\n① the task repeats weekly or more　② verification can be automated　③ the token budget can take it　④ the agent has tools that actually *run and see the result*\n\nMiss any one → **don't build a loop**; do it by hand or another way.\n> What stops most people isn't \"can I write a loop,\" it's \"does my repo deserve one.\" A repo that deserves a loop has a reconciliation baseline (golden sample \u002F upstream total) + tests + a lint guard. **A repo that doesn't deserve a loop will only have its errors amplified by one.**\n\n### Step 1 · Define a *machine-decidable* goal (the hard part — the loop lives or dies here)\n\nThe whole loop rides on the comparator's \"is it done yet?\" **The comparator can only work if your exit condition can be judged yes\u002Fno by a machine.**\n\n- Bad: Vague (\"make it good,\" \"write it sharper\") → the comparator can't judge → either it never passes (stuck retrying) or it guesses (passes\u002Fblocks at random).\n- Good: Decidable (\"all 96 unit tests green AND a change-list is produced,\" \"module-02 fields filled, pytest passes, business logic untouched\") → one check settles it; the loop converges cleanly.\n\n**Five-point goal framework:**\n1. **Done-criterion is machine-verifiable.**\n2. **Boundary conditions defined alongside the done-criterion** (\"what it must NOT do\") — anti-Goodhart; missing boundaries = a license to cheat.\n3. **Has a failure fallback** — retry cap N + escalate to a human when exceeded.\n4. **Goal is layered.**\n5. **Prefer reconciliation over assertion for the done-criterion** — anchor to external fact (golden sample \u002F upstream total \u002F financial tie-out \u002F platform back-office numbers) before your own assertions. \"All tests pass\" can be gamed (loosen asserts, fake mocks, swallow exceptions); \"diff vs the reference \u003C 0.01\" can't.\n\n> **Self-check:** read the goal to someone who doesn't know the domain — can they run one command and tell whether it's done? If not, it isn't decidable enough. Go back.\n\n### Step 2 · Pick the loop type\n\n| Your task | Loop type (cybernetic) | How it stops |\n|---|---|---|\n| Has a clear \"done\" test (write to done \u002F a batch of images processed) | **servo** (`\u002Fgoal`-style closed-loop) | stops on reaching the goal |\n| No endpoint, must keep maintaining a state (inventory alert \u002F scheduled health check) | **regulator** (`\u002Floop`-style thermostat) | never stops; acts only on change (dead-band suppresses noise) |\n| Periodic sampling, stop on a condition (watch a PR until CI is green) | **regulator with an exit** | stops when the exit condition holds |\n| Must \"ensure something happens on time\" | wrap the above in `\u002Fschedule` | cron fires it |\n\n> Rule of thumb: clear \"done\" test → servo; must keep maintaining, no endpoint → regulator; must \"happen on time\" → wrap a regulator in schedule.\n\n### Step 3 · Pick a skeleton\n\n**Maintenance type (tend something that exists) → document-driven dispatch.**\nThe loop isn't \"run a fixed check on a timer,\" it's **\"read a doc on a timer, and dispatch only when the doc changed.\"** The doc is the task queue + state machine + human interface.\nThree disciplines: ① the problem column is human-write-only, the result column is loop-write-only, **state advances one-way and never rolls back**; ② **the exit code is final** (if the script says exit 1, the script wins); ③ state advances only as far as \"awaiting verification\" — **the \"done\" cell is flipped by a human only.** The loop is the worker, not the acceptance officer.\n\n**Greenfield type (build from scratch) → plan \u002F build \u002F judge, three roles.**\n\n| Role | Does | Key |\n|---|---|---|\n| **Plan** | break the goal into a spec + **decidable acceptance conditions** | acceptance must be script-judgeable |\n| **Build** | write to the spec | **must not change the acceptance conditions** |\n| **Judge** | run acceptance **independently**; pass → stop, fail → return with the failure reason to Build | **independent + deterministic** |\n\nThree iron rules (all bet on the judge): ① **the judge must be independent** — not the same agent as Build (grading your own homework always inflates); ② **deterministic rules** — pytest \u002F reconciliation diff \u002F type check \u002F diff, never \"looks right\"; ③ **Build may not edit the acceptance conditions to pass**. Three failed retries → escalate to a human.\n\n### Step 4 · Add damping (against oscillation \u002F runaway)\n\nRetry cap, hard stop, human flips the last switch = damping. **Negative feedback with no damping oscillates** (the Ralph-Wiggum loop: spinning in place, burning tokens).\n\n### Step 5 · Land in three stages (don't go fully automatic on day one)\n\n① **Run it once by hand** (forces you to state exactly \"how the judge decides\") → ② harden into a skill \u002F Claude Code sub-agents (a main Claude loops, dispatching plan\u002Fbuild\u002Fjudge) → ③ hang it on cron for full automation.\n\n---\n\n## Action 2 — Review a loop (checklist = five failure modes)\n\n> Run the loop past each row. **Hitting any one = this loop will misfire; send it back.** These five are negative experience (gotchas) — worth more than positive rules.\n\n| # | Failure mode (how it breaks) | Review question (a hit = red) | Antibody |\n|---|---|---|---|\n| 1 | Goal is a correct platitude → **spins, burns money** | Can the exit condition be machine-judged yes\u002Fno? Or is it \"manage it well \u002F make it good\"? | Replace with a decidable result condition (Action 1·Step 1) |\n| 2 | \"Verification\" written as \"check if it looks ok\" → **agent confidently says fine and stops** | Is the judge the defendant itself? Does verification rest on \"looks right\" or deterministic rules? | Reconcile + exit code rules + independent judge |\n| 3 | (worst) Only gates on \"all tests pass\" → **agent deletes the tests** | Is there a boundary (\"what it must NOT do\")? Or only a done-criterion? | Done-criterion **+ boundary** together (the Goodhart antibody) |\n| 4 | Counts on the agent asking mid-run → **it won't; it runs the wrong answer to the end** | Is there any \"clarify only at runtime\" point? | **Front-load every clarification**; settle it once before launch |\n| 5 | Bloated CLAUDE.md + stale memory → **the faster it loops, the more it errs** | Are the docs\u002Fmemory it depends on fresh? Who maintains them? | Layered memory + periodic lint |\n\n**Plus three red lines (violate any = not allowed to go automatic):**\n- **Keep judgment with the human.** Acceptance \u002F the \"done\" cell is flipped by a human; the loop is not the acceptance officer.\n- **Responsibility doesn't transfer.** Anything whose failure you can't afford (merge the wrong PR \u002F publish the wrong thing \u002F misallocate money) → **don't hand over the authority automatically.**\n- **Counter-intuitive warning.** The more \"self-improving \u002F rewrites-its-own-rules\" a loop is, the **stricter the human review it needs** (to see what it rewrote the rules into) — not looser. The machine is too fast to intercept after the fact, so the human's judgment must sit **before the action** (a hard gate), not as a post-hoc patch.\n\n---\n\n## Worked example — reviewing a \"nightly green-keeper\" loop\n\nYou want a loop that runs every night and fixes whatever tests are failing.\n\n- **Naive goal:** \"make all tests pass.\" → Step-1 self-check fails: this is the bait for failure mode #3.\n- **Decidable goal (fixed):** \"all tests green **AND** no test file deleted or weakened **AND** coverage not lowered **AND** a change-list produced.\" Boundary now defined alongside the done-criterion.\n- **Type:** servo with a retry cap of 3 (Step 2 + Step 4).\n- **Skeleton:** plan\u002Fbuild\u002Fjudge — the **judge is CI run independently**, never the fixing agent (Step 3).\n\nNow run the **review checklist**, and it catches what the naive version would have missed:\n- **#3 hit** → the naive \"all tests pass\" lets the agent delete a failing test to \"win.\" Fixed by the boundary \"no test file deleted\u002Fweakened.\"\n- **#2 hit** → if the fixing agent also judged its own fix, it would pass itself. Fixed by \"judge = independent CI, deterministic.\"\n- **#4 hit** → if a fix is ambiguous, the agent won't stop to ask at 2 a.m.; it'll commit a guess. Fixed by front-loading: ambiguous fixes are left for the human, not guessed.\n- **Red line** → the loop opens a PR but **does not auto-merge**; the human flips the last switch (responsibility doesn't transfer).\n\nThe naive loop and the reviewed loop differ by four lines of constraint — and that's the difference between \"wakes you to a deleted test suite\" and \"wakes you to a clean PR.\"\n\n---\n\n## One-line close\n\n> The hard part of writing a loop isn't \"can I write a loop,\" it's **defining a goal a machine can reconcile** — decidable, bounded, reconciliation-based. The controller must be deterministic and external; keep judgment and the standard with the human; the system tends toward entropy, so maintain it.\n> **A loop only rewards someone who has already thought it through. Count on it to think for you, and it will happily think wrong, with you, at scale.**\n\n---\n\n> Lineage: Wiener's two-level feedback (*The Human Use of Human Beings*, 1950) for the judgment\u002Fexecution split and red lines; the plan\u002Fbuild\u002Fjudge pattern from Anatoli's *Loops explained* and Addy's *Loop Engineering*.\n> Mechanism layer (how to wire the loop architecture): see `autonomous-loops` \u002F `continuous-agent-loop`. This skill does not re-implement mechanism; it covers goal definition and runaway prevention only.\n",{"name":5,"metadata":24,"description":7},{"origin":25},"ECC","main","a73152352c0612252131e5cf84e91312a6cf17dc","utilities\u002Floop-design-check","https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs","https:\u002F\u002Fgithub.com\u002FHenrique1601","https:\u002F\u002Fbezerraportifolio.netlify.app\u002F","Full Stack Developer | React, Next.js, TypeScript, Node.js & IA | Construindo produtos digitais e soluções SaaS","2026-11-15 02:56:04",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add Henrique1601\u002FSkiilsToIAs --skill loop-design-check",{"tool":40,"command":41},"bunx","bunx skills add Henrique1601\u002FSkiilsToIAs --skill loop-design-check",{"tool":43,"command":44},"pnpm","pnpm dlx skills add Henrique1601\u002FSkiilsToIAs --skill loop-design-check","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs\u002Fblob\u002Fmain\u002Futilities\u002Floop-design-check\u002FSKILL.md",true,[],[49,56,63,71,78,85],{"skillNo":50,"name":51,"nameZh":6,"description":52,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":19},"01M7E2ANBW9V6AVX5JXNCEAWF8","rag-architect-agent","Agente Arquiteto de RAG, Embeddings & Busca Vetorial. Especialista em design e implementação de pipelines de Retrieval-Augmented Generation de alta fidelidade: chunking semântico, modelagem pgvector no PostgreSQL (Neon\u002FSupabase), busca híbrida (cosseno + BM25), reranking, validação via TypeSafe AI e filosofia Ponytail (anti-overengineering).","https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs\u002Fblob\u002Fmain\u002Fcustom-agents\u002Frag-architect-agent\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M7E29KKF8R02ZRGYE2MJM67G","rag-architecture-patterns","Padrões arquiteturais e implementação de pipelines RAG (Retrieval-Augmented Generation) de alta fidelidade para qualquer projeto ou sistema. Cobre chunking semântico, embeddings, busca híbrida (pgvector + BM25), reranking, validação via TypeSafe AI e filosofia Ponytail (anti-overengineering).","https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs\u002Fblob\u002Fmain\u002Fai-agents\u002Frag-architecture-patterns\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":67,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M7CWRA8E3BJ7CPY8N8EG28J1","skill-repair","Use this to fix and re-install agent skills that have failed installation.\nThis skill provides the necessary context and permissions to surgically update\nthe `manifest.json` after a fix has been applied.","Apache-2.0","https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs\u002Fblob\u002Fmain\u002Futilities\u002Fskill-repair\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},"01M7CWR66FZP8FA47WVZ4Z2EXC","security-scan","Scan your Claude Code configuration (.claude\u002F directory) for security vulnerabilities, misconfigurations, and injection risks using AgentShield. Checks CLAUDE.md, settings.json, MCP servers, hooks, and agent definitions.","https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs\u002Fblob\u002Fmain\u002Futilities\u002Fsecurity-scan\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},"01M7CWR0QPY58QPTR1P759TFNG","security-review","Use this skill when adding authentication, handling user input, working with secrets, creating API endpoints, or implementing payment\u002Fsensitive features. Provides comprehensive security checklist and patterns.","https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs\u002Fblob\u002Fmain\u002Futilities\u002Fsecurity-review\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},"01M7CWQXCE40XWT4FZ3VWZFGDP","security-bounty-hunter","Hunt for exploitable, bounty-worthy security issues in repositories. Focuses on remotely reachable vulnerabilities that qualify for real reports instead of noisy local-only findings.","https:\u002F\u002Fgithub.com\u002FHenrique1601\u002FSkiilsToIAs\u002Fblob\u002Fmain\u002Futilities\u002Fsecurity-bounty-hunter\u002FSKILL.md",[],[]]