[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AB4J7SMYS2D7RGAPHQBVEN":3,"skill-same:01M7AB4J7SMYS2D7RGAPHQBVEN":47},{"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":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":30,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M7AB4J7SMYS2D7RGAPHQBVEN","challenge",null,"Run a risk-scaled blind-spot discovery pass over a converged, exported frame BETWEEN `think` and `spec-to-plan` (the seventh origin skill, third leg in flow order): pressure-test the spec through structured lenses, route every finding back through the existing deterministic moves as proposed-only content the human adjudicates, and on a clean pass record the examined lenses\u002Fsurfaces and residual uncertainty — never a claim that there are no unknown unknowns. Use when the user says \"challenge this spec\", \"blind-spot pass\", \"pressure- test the frame\", \"what are we missing\", \"unknown unknowns\", or after `think` exports and before `devague plan new`. Authored and maintained in agentculture\u002Fdevague (origin = devague); guildmaster pulls this skill from here and broadcasts it to the AgentCulture mesh — it is NOT vendored from guildmaster like the inbound skills here.","agentculture\u002Fculture-nodes","agentculture","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F277824871?v=4","01M7AAZDPCFA8RDCA27DQKKWCJ","AgentCulture","01M7AAZDQ1B6FYE24021YCDAPR",3,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes\u002Fblob\u002Fmain\u002F.qwen\u002Fskills\u002Fchallenge\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: challenge\ndescription: >-\n  Run a risk-scaled blind-spot discovery pass over a converged, exported frame\n  BETWEEN `think` and `spec-to-plan` (the seventh origin skill, third leg in\n  flow order): pressure-test the spec through structured lenses, route every\n  finding back through the existing deterministic moves as proposed-only content\n  the human adjudicates, and on a clean pass record the examined lenses\u002Fsurfaces\n  and residual uncertainty — never a claim that there are no unknown unknowns.\n  Use when the user says \"challenge this spec\", \"blind-spot pass\", \"pressure-\n  test the frame\", \"what are we missing\", \"unknown unknowns\", or after `think`\n  exports and before `devague plan new`. Authored and maintained in\n  agentculture\u002Fdevague (origin = devague); guildmaster pulls this skill from\n  here and broadcasts it to the AgentCulture mesh — it is NOT vendored from\n  guildmaster like the inbound skills here.\n---\n\n\u003C!-- lineage: claude (.claude\u002Fskills\u002Fchallenge\u002FSKILL.md) -> colleague (..\u002Fcolleague\u002F.colleague\u002Fskills\u002Fchallenge.md, ff7331e v1.63.1) -> qwen (this file). Surface adaptation: qwen drives the devague CLI directly through its shell tool; there is no curated tool allow-list to navigate. -->\n\n# challenge — hunt the blind spots before the plan inherits them\n\nThe skill is named **`challenge`**; it is the **blind-spot discovery leg** of\nthe devague method — the *seventh* origin skill, sitting *third* in flow\norder, between the spec leg and the plan leg:\n\n```text\nscope -> think -> challenge -> spec-to-plan -> assign-to-workforce -> deviate -> summarize-delivery\n```\n\nBefore this leg existed, a frame could converge on precisely stated claims\nwhile the original framing was still incomplete: open questions only captured\nuncertainty someone had already noticed — **nothing actively hunted omitted\ndimensions, hidden dependencies, or assumptions shared by everyone in the\nframe, and no record existed of which surfaces were ever examined**\n(issue 73's problem statement). Strictly speaking, an unknown unknown cannot\nbe listed directly — once articulated, it becomes a known unknown. The useful\ncapability is therefore to **raise the odds of discovering blind spots and\nlower the cost of the surprises that remain**, not to promise their\nelimination.\n\nThat is the surprise-cost rationale: an articulated blind spot becomes a\nknown unknown the method can manage; an unexamined one surfaces later as a\nmid-run `deviate` or a production surprise. Discovery *before* planning is\ncheaper than either — a proposed claim the human rejects costs minutes; the\nsame gap found mid-fan-out stops a wave, and found in production it costs\nwhatever the blast radius costs.\n\nThis doc is written for two readers. The **operator** — the main agent — runs\nthe pass: sweeps the lenses, drives the deterministic CLI move by move, and\nproposes findings. The **gate-owning human** adjudicates: every finding lands\n`proposed`, and confirming, rejecting, or resolving it is the human exercising\nthe existing **spec gate** (gate 1) — challenge adds no fourth gate, mirroring\nhow `deviate` amends gate 2 rather than adding one.\n\n## How to run in qwen\n\nThe pass drives the `devague` CLI — the same agent-agnostic commands for any\nagent or operator (if `devague` isn't on `PATH`: `uv tool install devague`).\nQwen Code has no curated tool allow-list: every move — `capture`,\n`interrogate`, `park`, `converge`, `scope`, `question`, `review`, `export`,\nand `plan risk` — runs directly through the shell tool\n(`run_shell_command`), and the operator runs the same commands directly in a\nterminal. `confirm` and `reject` are user-only decisions either way.\n\n## When it runs\n\nThe timing is a recorded decision — quote it, don't re-derive it:\n\n> the challenge pass runs after \u002Fthink exports: challenge the converged,\n> exported frame before `devague plan new`; findings reopen the frame,\n> reconverge, and re-export the same dated spec file — \u002Fthink stays\n> self-contained (resolves q1)\n\n— decision c17 in `docs\u002Fspecs\u002F2026-07-15-challenge-skill.md`.\n\nConcretely: `think` finishes its own arc (converge, export) untouched. Then\n`challenge` pressure-tests the exported spec **before** `devague plan new`\nseeds a plan from it. Findings land as proposed claims, honesty conditions,\nquestions, or parks — which reopens the frame — the human adjudicates, the\nframe **reconverges**, and `devague export` **re-exports the same dated\nfile** (`docs\u002Fspecs\u002F\u003Ccreated-date>-\u003Cslug>.md`; exports are prefixed with the\nframe's creation date, so a re-export overwrites in place rather than\nspawning a duplicate). That reconverge-and-re-export loop repeats until the\npass's findings are all adjudicated and the spec artifact carries them.\n\n## Proportionality — scale the pass to the risk\n\nThe pass is **mandatory but proportional**: lightweight for ordinary work,\nrigorous for high-risk work (integration option 1 from issue 73, per the\nissue author — a distinct operator skill, no new CLI engine until the\nworkflow proves it needs one). Which work is high-risk is likewise a recorded\ndecision:\n\n> the named escalation signals that deepen the pass from lightweight to\n> rigorous: migrations, security-sensitive work, distributed state, hardware,\n> destructive operations, other hard-to-reverse changes, concurrency hazards,\n> and any surface that can lose user data (resolves q3)\n\n— decision c19 in `docs\u002Fspecs\u002F2026-07-15-challenge-skill.md`.\n\nIf **any** escalation signal applies, run the rigorous form: every lens,\ndeliberate counter-evidence hunting, and cheap probes where they would settle\na real question. If none applies, a lightweight sweep — one pass over the\nlenses against the exported spec, minutes not hours — satisfies the method.\nLightweight never means skipped: even the lightest pass leaves durable\nrecords (see the hard rules).\n\n## The lenses\n\nSweep the exported spec, the live frame, and the surfaces the idea touches\nthrough these structured lenses (from issue 73):\n\n- **adjacent systems and hidden dependencies** — what else reads, writes, or\n  assumes the thing being changed;\n- **unstated assumptions and missing counter-evidence** — what everyone in\n  the frame believes without a claim saying so, and what was never checked\n  because nobody argued the other side;\n- **overlooked actors, lifecycle stages, data flows, and failure modes** —\n  who and what the frame forgot: other users, upgrade\u002Fdowngrade paths,\n  half-completed operations;\n- **security, migration, concurrency, operations, and reversibility** — the\n  classic hard surfaces, each also an escalation signal when present;\n- **missing observability, containment, rollback, and recovery paths** —\n  when the surprise happens anyway, how it is seen, bounded, and undone;\n- **cheap probes or experiments** — small, scratch-space checks that could\n  expose a surprise now instead of mid-run (probes never mutate the repo or\n  `.devague\u002F` state).\n\n## The method\n\n1. **Confirm the entry condition.** A converged frame that `think` has\n   already exported, and no plan seeded from it yet (`devague status` shows\n   where the frame stands; the exported spec-md is the artifact under\n   challenge).\n2. **Set the depth.** Check the idea against the c19 escalation signals\n   above. Any hit → rigorous; none → lightweight. Say which you chose and\n   why — the depth decision is part of the pass's record.\n3. **Sweep the lenses read-only.** Read the exported spec claim by claim,\n   the frame's parked vagueness and resolved questions, and the actual\n   surfaces the idea touches. Nothing mutates during the sweep; the only\n   mutations are the deterministic moves that record findings.\n4. **Route every finding through an existing move.** Use the routing table\n   below. Everything the agent proposes carries `--origin llm` and lands\n   `proposed` — the pass cannot silently convert speculation into confirmed\n   requirements.\n5. **Let the human adjudicate.** `devague review` lists every proposal with\n   ids; `devague confirm` \u002F `devague reject` \u002F `devague question --resolve`\n   are user-only decisions. This is the existing spec gate doing its job.\n6. **Reconverge and re-export.** `devague converge`, then `devague export` —\n   the same dated spec file now carries the adjudicated findings and the\n   pass's provenance (the exported spec renders scope entries in its\n   Scope exploration section).\n7. **On a clean pass, record the pass itself.** A sweep that finds nothing\n   still records which lenses and surfaces were examined (`devague scope`\n   entries, one per lens\u002Fsurface, e.g. `challenge pass \u002F concurrency lens:\n   devague\u002Fstore.py`) and what residual uncertainty remains (`park`). A bare\n   \"no issues found\" is not an outcome this skill produces.\n\n## Where findings land\n\nChallenge keeps **no parallel prose-only artifact** — the frame is the\nrecord, and every output category from issue 73 has an existing deterministic\nmove to land in:\n\n| Output category (issue 73) | What it is | Landing move |\n|----------------------------|------------|--------------|\n| known facts | something the pass established, with provenance | `capture --kind requirement` \u002F `--kind decision` \u002F `--kind boundary` (`--origin llm` → lands `proposed`) |\n| assumptions | beliefs the frame leaned on unstated | `capture --kind assumption --origin llm`, then pressure-test with `interrogate --honesty` \u002F `--hard-question` \u002F `--contradicts` |\n| known unknowns \u002F open questions | articulated uncertainty | `question \"\u003Ctext>\"` when it needs a user decision; `park --kind unknown_nonblocking\\|unknown_blocking` when not decidable now |\n| unexamined surfaces | what this pass did not (or could not) look at | `devague scope \"\u003Csurface>\" --finding \"\u003Cwhat was and wasn't examined, and why>\"` |\n| residual surprise risk | uncertainty that survives the pass | `park` on the frame while speccing; `devague plan risk --kind \u003Ckind>` once the plan exists |\n| resilience measures | containment, rollback, recovery the surprise cost demands | spec-side `capture --kind requirement` \u002F `--kind boundary`; plan-side `devague plan risk` (see below) |\n\nEvery finding names the **lens and surface** it came from (the\n`challenge pass \u002F \u003Clens>: \u003Csurface>` convention in scope entries; provenance\ncitations inside claim text). That provenance bar is how the method hunts\nblind spots **without encouraging speculative issue generation** — a finding\nyou cannot trace to something you actually read is speculation, not a\nfinding.\n\n## Resilience placement — spec-side or plan-side\n\nWhere a resilience measure lands is a recorded decision:\n\n> resilience measures land in both spec and plan by nature: spec-side as\n> requirement\u002Fboundary claims when they change what to build, plan-side as\n> plan risks or tasks when they change how to build it — the skill coaches\n> which is which (resolves q2)\n\n— decision c18 in `docs\u002Fspecs\u002F2026-07-15-challenge-skill.md`.\n\nThe coaching: ask *\"does this change **what** ships, or **how** it gets\nbuilt?\"* A rollback path the user needs, a fail-closed version check, a\ncontainment boundary — those change the product: `capture` them spec-side as\n`requirement` \u002F `boundary` claims so the re-exported spec carries them.\nA staging sequence, a merge-order constraint, an uncertainty the workforce\nmust build around — those change the build: land them plan-side via\n`devague plan risk --kind \u003Ckind>` (blocking or nonblocking, honestly chosen)\nonce `spec-to-plan` seeds the plan, where the plan's convergence gate keeps\nblocking risks visible until resolved.\n\n## Hard rules (do not violate)\n\n- **Never conclude there are no unknown unknowns.** Not after a rigorous\n  pass, not after a clean one. The only honest clean-pass output is a record\n  of which lenses and surfaces were examined — `devague scope` entries — plus\n  the residual uncertainty that remains — `park` — so the pass leaves durable\n  provenance instead of a comforting absolute (issue 73 success criteria; the\n  anti-fabrication contract in `docs\u002Fllm-guidance.md`).\n- **LLM-origin findings stay proposed until the user confirms.** Every\n  finding the agent proposes carries `--origin llm` and lands `proposed`;\n  only the user's `confirm` makes it real. The pass must not be able to\n  silently convert speculation into confirmed requirements.\n- **Findings route through existing deterministic moves only.** `capture`,\n  `interrogate`, `question`, `park`, `devague scope`, `devague plan risk` —\n  nothing else. No parallel prose artifact, no new CLI verb, engine, or\n  state model (issue 20; issue 73's stated preference). If it didn't land in\n  a move, it didn't land.\n- **Provenance on every finding.** Name the lens and the surface it came\n  from. If you didn't read it, don't claim it — same bar as `scope`.\n- **Proportional, never skipped.** Lightweight is the floor, not an\n  exemption; any c19 escalation signal makes the rigorous form mandatory.\n  Skipping the pass entirely is not a depth setting.\n- **Not a fourth standing gate.** The three human gates stay: exported spec,\n  implementation split plan, final PR. Challenge output is adjudicated\n  inside the existing spec gate — mirroring how `deviate` amends gate 2\n  rather than adding one.\n- **The sweep is read-only until it routes.** Reading spec, frame, and\n  surfaces never edits files or state; probes run in scratch space. The only\n  mutations are the recording moves themselves.\n\n## Worked example\n\nChallenging the exported spec for a store-schema migration (illustrative\nslug `store-schema-v3`) — \"migrations\" and \"any surface that can lose user\ndata\" are both c19 escalation signals, so the pass runs rigorous:\n\n```bash\n# Entry condition: think exported docs\u002Fspecs\u002F2026-07-15-store-schema-v3.md\n# and no plan exists yet. Depth: rigorous (migration + data-loss signals).\n\n# adjacent-systems lens: an older installed devague reads the same store\ndevague capture --origin llm --kind assumption \"older installed devague binaries refuse a v3 store via the fail-closed schema_version check in devague\u002Fstore.py\"\ndevague interrogate c9 --origin llm --honesty \"a v2-reading binary pointed at a v3 store exits with the version hint, not a traceback\"\ndevague scope \"challenge pass \u002F adjacent-systems lens: devague\u002Fstore.py schema_version gate\" --finding \"older binaries fail closed on v3; seeded the compat assumption\" --seeds c9\n\n# failure-mode lens: the migration can die halfway\ndevague capture --origin llm --kind requirement \"migration writes to a temp file and renames — a killed run never leaves a half-written store\"\n\n# overlooked-actors lens: needs a user decision, not a guess\ndevague question \"do mesh agents share one store, or does each checkout own its own?\"\n\n# reversibility lens: genuinely unknown, not decidable now\ndevague park \"whether a v3->v2 downgrade path is ever needed\" --kind unknown_nonblocking\n\n# concurrency lens found nothing — record the clean pass, not a conclusion\ndevague scope \"challenge pass \u002F concurrency lens: devague\u002Fstore.py + delivery_store.py\" --finding \"single-writer CLI, no locking today; clean pass — residual risk only if two agents ever share a checkout\"\n\n# --- HUMAN adjudicates: the existing spec gate at work ---\ndevague review\ndevague confirm c9 h4 c10\ndevague question --resolve q1 --decision \"each checkout owns its own store\"\n\n# Reconverge and re-export — the SAME dated file, now carrying the pass\ndevague converge\ndevague export\n\n# Residual risk that changes HOW to build lands plan-side once\n# spec-to-plan seeds the plan:\ndevague plan new --frame store-schema-v3\ndevague plan risk \"two agents sharing a checkout could interleave store writes mid-migration\" --kind unknown_nonblocking\n```\n\nEvery finding above is traceable to a lens and a surface; the clean lens is\nrecorded as examined rather than silently dropped; and nothing the agent\nproposed became confirmed without the human's `confirm`.\n\n## After the pass — hand off to spec-to-plan\n\nOnce the frame reconverges and the same dated spec file is re-exported, the\npass is done — the examined surfaces, residual uncertainty, and adjudicated\nfindings all live in frame state and render into the spec artifact. Continue\nwith `spec-to-plan` as usual: the plan seeds from the challenged frame, and\nany residual surprise risk you routed plan-side lands via\n`devague plan risk` as first-class plan state. If a surprise still gets\nthrough mid-fan-out, that is `deviate`'s job — and every approved `dN`\ndeviation record is evidence for what the *next* challenge pass's lenses\nshould look harder at.\n\n## Provenance\n\nThis is a **first-party** skill — its origin is `agentculture\u002Fdevague`, the\n*seventh* in the outbound family after `scope`, `think`, `spec-to-plan`,\n`assign-to-workforce`, `deviate`, and `summarize-delivery`, sitting third\nin flow order as the blind-spot discovery leg between `think` and\n`spec-to-plan`. guildmaster pulls it from here and broadcasts it to the\nAgentCulture mesh; because devague is upstream, it is **never re-vendored\nback** from guildmaster's re-broadcast copy. The `cite, don't import` policy\nstill holds: downstream repos copy it, they don't symlink or depend on it.\nSee `docs\u002Fskill-sources.md`.\n","\u003C!-- lineage: claude (.claude\u002Fskills\u002Fchallenge\u002FSKILL.md) -> colleague (..\u002Fcolleague\u002F.colleague\u002Fskills\u002Fchallenge.md, ff7331e v1.63.1) -> qwen (this file). Surface adaptation: qwen drives the devague CLI directly through its shell tool; there is no curated tool allow-list to navigate. -->\n\n# challenge — hunt the blind spots before the plan inherits them\n\nThe skill is named **`challenge`**; it is the **blind-spot discovery leg** of\nthe devague method — the *seventh* origin skill, sitting *third* in flow\norder, between the spec leg and the plan leg:\n\n```text\nscope -> think -> challenge -> spec-to-plan -> assign-to-workforce -> deviate -> summarize-delivery\n```\n\nBefore this leg existed, a frame could converge on precisely stated claims\nwhile the original framing was still incomplete: open questions only captured\nuncertainty someone had already noticed — **nothing actively hunted omitted\ndimensions, hidden dependencies, or assumptions shared by everyone in the\nframe, and no record existed of which surfaces were ever examined**\n(issue 73's problem statement). Strictly speaking, an unknown unknown cannot\nbe listed directly — once articulated, it becomes a known unknown. The useful\ncapability is therefore to **raise the odds of discovering blind spots and\nlower the cost of the surprises that remain**, not to promise their\nelimination.\n\nThat is the surprise-cost rationale: an articulated blind spot becomes a\nknown unknown the method can manage; an unexamined one surfaces later as a\nmid-run `deviate` or a production surprise. Discovery *before* planning is\ncheaper than either — a proposed claim the human rejects costs minutes; the\nsame gap found mid-fan-out stops a wave, and found in production it costs\nwhatever the blast radius costs.\n\nThis doc is written for two readers. The **operator** — the main agent — runs\nthe pass: sweeps the lenses, drives the deterministic CLI move by move, and\nproposes findings. The **gate-owning human** adjudicates: every finding lands\n`proposed`, and confirming, rejecting, or resolving it is the human exercising\nthe existing **spec gate** (gate 1) — challenge adds no fourth gate, mirroring\nhow `deviate` amends gate 2 rather than adding one.\n\n## How to run in qwen\n\nThe pass drives the `devague` CLI — the same agent-agnostic commands for any\nagent or operator (if `devague` isn't on `PATH`: `uv tool install devague`).\nQwen Code has no curated tool allow-list: every move — `capture`,\n`interrogate`, `park`, `converge`, `scope`, `question`, `review`, `export`,\nand `plan risk` — runs directly through the shell tool\n(`run_shell_command`), and the operator runs the same commands directly in a\nterminal. `confirm` and `reject` are user-only decisions either way.\n\n## When it runs\n\nThe timing is a recorded decision — quote it, don't re-derive it:\n\n> the challenge pass runs after \u002Fthink exports: challenge the converged,\n> exported frame before `devague plan new`; findings reopen the frame,\n> reconverge, and re-export the same dated spec file — \u002Fthink stays\n> self-contained (resolves q1)\n\n— decision c17 in `docs\u002Fspecs\u002F2026-07-15-challenge-skill.md`.\n\nConcretely: `think` finishes its own arc (converge, export) untouched. Then\n`challenge` pressure-tests the exported spec **before** `devague plan new`\nseeds a plan from it. Findings land as proposed claims, honesty conditions,\nquestions, or parks — which reopens the frame — the human adjudicates, the\nframe **reconverges**, and `devague export` **re-exports the same dated\nfile** (`docs\u002Fspecs\u002F\u003Ccreated-date>-\u003Cslug>.md`; exports are prefixed with the\nframe's creation date, so a re-export overwrites in place rather than\nspawning a duplicate). That reconverge-and-re-export loop repeats until the\npass's findings are all adjudicated and the spec artifact carries them.\n\n## Proportionality — scale the pass to the risk\n\nThe pass is **mandatory but proportional**: lightweight for ordinary work,\nrigorous for high-risk work (integration option 1 from issue 73, per the\nissue author — a distinct operator skill, no new CLI engine until the\nworkflow proves it needs one). Which work is high-risk is likewise a recorded\ndecision:\n\n> the named escalation signals that deepen the pass from lightweight to\n> rigorous: migrations, security-sensitive work, distributed state, hardware,\n> destructive operations, other hard-to-reverse changes, concurrency hazards,\n> and any surface that can lose user data (resolves q3)\n\n— decision c19 in `docs\u002Fspecs\u002F2026-07-15-challenge-skill.md`.\n\nIf **any** escalation signal applies, run the rigorous form: every lens,\ndeliberate counter-evidence hunting, and cheap probes where they would settle\na real question. If none applies, a lightweight sweep — one pass over the\nlenses against the exported spec, minutes not hours — satisfies the method.\nLightweight never means skipped: even the lightest pass leaves durable\nrecords (see the hard rules).\n\n## The lenses\n\nSweep the exported spec, the live frame, and the surfaces the idea touches\nthrough these structured lenses (from issue 73):\n\n- **adjacent systems and hidden dependencies** — what else reads, writes, or\n  assumes the thing being changed;\n- **unstated assumptions and missing counter-evidence** — what everyone in\n  the frame believes without a claim saying so, and what was never checked\n  because nobody argued the other side;\n- **overlooked actors, lifecycle stages, data flows, and failure modes** —\n  who and what the frame forgot: other users, upgrade\u002Fdowngrade paths,\n  half-completed operations;\n- **security, migration, concurrency, operations, and reversibility** — the\n  classic hard surfaces, each also an escalation signal when present;\n- **missing observability, containment, rollback, and recovery paths** —\n  when the surprise happens anyway, how it is seen, bounded, and undone;\n- **cheap probes or experiments** — small, scratch-space checks that could\n  expose a surprise now instead of mid-run (probes never mutate the repo or\n  `.devague\u002F` state).\n\n## The method\n\n1. **Confirm the entry condition.** A converged frame that `think` has\n   already exported, and no plan seeded from it yet (`devague status` shows\n   where the frame stands; the exported spec-md is the artifact under\n   challenge).\n2. **Set the depth.** Check the idea against the c19 escalation signals\n   above. Any hit → rigorous; none → lightweight. Say which you chose and\n   why — the depth decision is part of the pass's record.\n3. **Sweep the lenses read-only.** Read the exported spec claim by claim,\n   the frame's parked vagueness and resolved questions, and the actual\n   surfaces the idea touches. Nothing mutates during the sweep; the only\n   mutations are the deterministic moves that record findings.\n4. **Route every finding through an existing move.** Use the routing table\n   below. Everything the agent proposes carries `--origin llm` and lands\n   `proposed` — the pass cannot silently convert speculation into confirmed\n   requirements.\n5. **Let the human adjudicate.** `devague review` lists every proposal with\n   ids; `devague confirm` \u002F `devague reject` \u002F `devague question --resolve`\n   are user-only decisions. This is the existing spec gate doing its job.\n6. **Reconverge and re-export.** `devague converge`, then `devague export` —\n   the same dated spec file now carries the adjudicated findings and the\n   pass's provenance (the exported spec renders scope entries in its\n   Scope exploration section).\n7. **On a clean pass, record the pass itself.** A sweep that finds nothing\n   still records which lenses and surfaces were examined (`devague scope`\n   entries, one per lens\u002Fsurface, e.g. `challenge pass \u002F concurrency lens:\n   devague\u002Fstore.py`) and what residual uncertainty remains (`park`). A bare\n   \"no issues found\" is not an outcome this skill produces.\n\n## Where findings land\n\nChallenge keeps **no parallel prose-only artifact** — the frame is the\nrecord, and every output category from issue 73 has an existing deterministic\nmove to land in:\n\n| Output category (issue 73) | What it is | Landing move |\n|----------------------------|------------|--------------|\n| known facts | something the pass established, with provenance | `capture --kind requirement` \u002F `--kind decision` \u002F `--kind boundary` (`--origin llm` → lands `proposed`) |\n| assumptions | beliefs the frame leaned on unstated | `capture --kind assumption --origin llm`, then pressure-test with `interrogate --honesty` \u002F `--hard-question` \u002F `--contradicts` |\n| known unknowns \u002F open questions | articulated uncertainty | `question \"\u003Ctext>\"` when it needs a user decision; `park --kind unknown_nonblocking\\|unknown_blocking` when not decidable now |\n| unexamined surfaces | what this pass did not (or could not) look at | `devague scope \"\u003Csurface>\" --finding \"\u003Cwhat was and wasn't examined, and why>\"` |\n| residual surprise risk | uncertainty that survives the pass | `park` on the frame while speccing; `devague plan risk --kind \u003Ckind>` once the plan exists |\n| resilience measures | containment, rollback, recovery the surprise cost demands | spec-side `capture --kind requirement` \u002F `--kind boundary`; plan-side `devague plan risk` (see below) |\n\nEvery finding names the **lens and surface** it came from (the\n`challenge pass \u002F \u003Clens>: \u003Csurface>` convention in scope entries; provenance\ncitations inside claim text). That provenance bar is how the method hunts\nblind spots **without encouraging speculative issue generation** — a finding\nyou cannot trace to something you actually read is speculation, not a\nfinding.\n\n## Resilience placement — spec-side or plan-side\n\nWhere a resilience measure lands is a recorded decision:\n\n> resilience measures land in both spec and plan by nature: spec-side as\n> requirement\u002Fboundary claims when they change what to build, plan-side as\n> plan risks or tasks when they change how to build it — the skill coaches\n> which is which (resolves q2)\n\n— decision c18 in `docs\u002Fspecs\u002F2026-07-15-challenge-skill.md`.\n\nThe coaching: ask *\"does this change **what** ships, or **how** it gets\nbuilt?\"* A rollback path the user needs, a fail-closed version check, a\ncontainment boundary — those change the product: `capture` them spec-side as\n`requirement` \u002F `boundary` claims so the re-exported spec carries them.\nA staging sequence, a merge-order constraint, an uncertainty the workforce\nmust build around — those change the build: land them plan-side via\n`devague plan risk --kind \u003Ckind>` (blocking or nonblocking, honestly chosen)\nonce `spec-to-plan` seeds the plan, where the plan's convergence gate keeps\nblocking risks visible until resolved.\n\n## Hard rules (do not violate)\n\n- **Never conclude there are no unknown unknowns.** Not after a rigorous\n  pass, not after a clean one. The only honest clean-pass output is a record\n  of which lenses and surfaces were examined — `devague scope` entries — plus\n  the residual uncertainty that remains — `park` — so the pass leaves durable\n  provenance instead of a comforting absolute (issue 73 success criteria; the\n  anti-fabrication contract in `docs\u002Fllm-guidance.md`).\n- **LLM-origin findings stay proposed until the user confirms.** Every\n  finding the agent proposes carries `--origin llm` and lands `proposed`;\n  only the user's `confirm` makes it real. The pass must not be able to\n  silently convert speculation into confirmed requirements.\n- **Findings route through existing deterministic moves only.** `capture`,\n  `interrogate`, `question`, `park`, `devague scope`, `devague plan risk` —\n  nothing else. No parallel prose artifact, no new CLI verb, engine, or\n  state model (issue 20; issue 73's stated preference). If it didn't land in\n  a move, it didn't land.\n- **Provenance on every finding.** Name the lens and the surface it came\n  from. If you didn't read it, don't claim it — same bar as `scope`.\n- **Proportional, never skipped.** Lightweight is the floor, not an\n  exemption; any c19 escalation signal makes the rigorous form mandatory.\n  Skipping the pass entirely is not a depth setting.\n- **Not a fourth standing gate.** The three human gates stay: exported spec,\n  implementation split plan, final PR. Challenge output is adjudicated\n  inside the existing spec gate — mirroring how `deviate` amends gate 2\n  rather than adding one.\n- **The sweep is read-only until it routes.** Reading spec, frame, and\n  surfaces never edits files or state; probes run in scratch space. The only\n  mutations are the recording moves themselves.\n\n## Worked example\n\nChallenging the exported spec for a store-schema migration (illustrative\nslug `store-schema-v3`) — \"migrations\" and \"any surface that can lose user\ndata\" are both c19 escalation signals, so the pass runs rigorous:\n\n```bash\n# Entry condition: think exported docs\u002Fspecs\u002F2026-07-15-store-schema-v3.md\n# and no plan exists yet. Depth: rigorous (migration + data-loss signals).\n\n# adjacent-systems lens: an older installed devague reads the same store\ndevague capture --origin llm --kind assumption \"older installed devague binaries refuse a v3 store via the fail-closed schema_version check in devague\u002Fstore.py\"\ndevague interrogate c9 --origin llm --honesty \"a v2-reading binary pointed at a v3 store exits with the version hint, not a traceback\"\ndevague scope \"challenge pass \u002F adjacent-systems lens: devague\u002Fstore.py schema_version gate\" --finding \"older binaries fail closed on v3; seeded the compat assumption\" --seeds c9\n\n# failure-mode lens: the migration can die halfway\ndevague capture --origin llm --kind requirement \"migration writes to a temp file and renames — a killed run never leaves a half-written store\"\n\n# overlooked-actors lens: needs a user decision, not a guess\ndevague question \"do mesh agents share one store, or does each checkout own its own?\"\n\n# reversibility lens: genuinely unknown, not decidable now\ndevague park \"whether a v3->v2 downgrade path is ever needed\" --kind unknown_nonblocking\n\n# concurrency lens found nothing — record the clean pass, not a conclusion\ndevague scope \"challenge pass \u002F concurrency lens: devague\u002Fstore.py + delivery_store.py\" --finding \"single-writer CLI, no locking today; clean pass — residual risk only if two agents ever share a checkout\"\n\n# --- HUMAN adjudicates: the existing spec gate at work ---\ndevague review\ndevague confirm c9 h4 c10\ndevague question --resolve q1 --decision \"each checkout owns its own store\"\n\n# Reconverge and re-export — the SAME dated file, now carrying the pass\ndevague converge\ndevague export\n\n# Residual risk that changes HOW to build lands plan-side once\n# spec-to-plan seeds the plan:\ndevague plan new --frame store-schema-v3\ndevague plan risk \"two agents sharing a checkout could interleave store writes mid-migration\" --kind unknown_nonblocking\n```\n\nEvery finding above is traceable to a lens and a surface; the clean lens is\nrecorded as examined rather than silently dropped; and nothing the agent\nproposed became confirmed without the human's `confirm`.\n\n## After the pass — hand off to spec-to-plan\n\nOnce the frame reconverges and the same dated spec file is re-exported, the\npass is done — the examined surfaces, residual uncertainty, and adjudicated\nfindings all live in frame state and render into the spec artifact. Continue\nwith `spec-to-plan` as usual: the plan seeds from the challenged frame, and\nany residual surprise risk you routed plan-side lands via\n`devague plan risk` as first-class plan state. If a surprise still gets\nthrough mid-fan-out, that is `deviate`'s job — and every approved `dN`\ndeviation record is evidence for what the *next* challenge pass's lenses\nshould look harder at.\n\n## Provenance\n\nThis is a **first-party** skill — its origin is `agentculture\u002Fdevague`, the\n*seventh* in the outbound family after `scope`, `think`, `spec-to-plan`,\n`assign-to-workforce`, `deviate`, and `summarize-delivery`, sitting third\nin flow order as the blind-spot discovery leg between `think` and\n`spec-to-plan`. guildmaster pulls it from here and broadcasts it to the\nAgentCulture mesh; because devague is upstream, it is **never re-vendored\nback** from guildmaster's re-broadcast copy. The `cite, don't import` policy\nstill holds: downstream repos copy it, they don't symlink or depend on it.\nSee `docs\u002Fskill-sources.md`.\n",{"name":5,"description":7},"main","a2056e6191a4e106faf886773efed2f44ff120b2",".qwen\u002Fskills\u002Fchallenge",0,"https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes","https:\u002F\u002Fgithub.com\u002Fagentculture","https:\u002F\u002Fagentculture.org","Agent first OSS software","2026-09-29 08:50:58",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add agentculture\u002Fculture-nodes --skill challenge",{"tool":39,"command":40},"bunx","bunx skills add agentculture\u002Fculture-nodes --skill challenge",{"tool":42,"command":43},"pnpm","pnpm dlx skills add agentculture\u002Fculture-nodes --skill challenge","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes\u002Fblob\u002Fmain\u002F.qwen\u002Fskills\u002Fchallenge\u002FSKILL.md",true,[],[48,55,60,65,72,79],{"skillNo":49,"name":50,"nameZh":6,"description":51,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":52,"categories":53,"keywords":54,"publishedAt":19},"01M7AB6FGWR0P3SPYFX2W801D5","sample","Fixture skill proving the .qwen\u002Fskills leg carries at least one skill.","https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes\u002Fblob\u002Fmain\u002Fadapters\u002Fqwen\u002Ftests\u002Ffixtures\u002Fconfig-repo-no-culture-yaml\u002F.qwen\u002Fskills\u002Fsample\u002FSKILL.md",[],[],{"skillNo":56,"name":50,"nameZh":6,"description":51,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":19},"01M7AB6E9PFCDZE193CGPM0K35","https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes\u002Fblob\u002Fmain\u002Fadapters\u002Fqwen\u002Ftests\u002Ffixtures\u002Fconfig-repo-no-agents-md\u002F.qwen\u002Fskills\u002Fsample\u002FSKILL.md",[],[],{"skillNo":61,"name":50,"nameZh":6,"description":51,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":19},"01M7AB6CJ3G1ZZJD63NEZZK1JH","https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes\u002Fblob\u002Fmain\u002Fadapters\u002Fqwen\u002Ftests\u002Ffixtures\u002Fconfig-repo-complete\u002F.qwen\u002Fskills\u002Fsample\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":19},"01M7AB64WWKV9H2J7JDV5KXGPA","think","Think a vague feature idea into a buildable spec by working backwards (the idea→spec leg; drives the `devague` CLI). Start from the announcement (\"pretend it shipped\"), capture and classify claims, interrogate them with honesty conditions and hard questions, park open vagueness as a first-class object, and export a spec only once the frame *converges*. Use when the user says \"think this through\", \"spec this\", \"work backwards\", \"turn this idea into a spec\", \"announcement frame\", or \"devague\", or when a feature request is too vague to build yet. Explore scope first when the idea touches an existing codebase — the sibling `scope` skill is the optional opening leg ahead of `think`. Once a spec exports, hand off to the sibling `spec-to-plan` skill to turn it into a plan. Authored and maintained in agentculture\u002Fdevague (origin = devague); guildmaster pulls this skill from here and broadcasts it to the AgentCulture mesh — it is NOT vendored from guildmaster like the other skills here.","https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes\u002Fblob\u002Fmain\u002F.qwen\u002Fskills\u002Fthink\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":19},"01M7AB5X0FTZ081AFH3BTGQKVA","summarize-delivery","Close the loop after an `assign-to-workforce` run by turning what actually happened into an accountability artifact — planned versus actual delivery, mid-work decisions, plan drift, evidence-backed delivery claims, and remaining work. The plan the user confirmed is the contract; this skill records where execution obeyed it, where it changed, and what is genuinely safe to claim as delivered. Runs on complete, partial, AND failed runs — failure is reported faithfully, never smoothed over. Use when the user says \"summarize delivery\", \"delivery summary\", \"wrap up\", \"close the loop\", \"what did we actually ship\", or \"plan versus actual\", or after `assign-to-workforce` merges (or fails to merge) a plan's waves. Authored and maintained in agentculture\u002Fdevague (origin = devague); guildmaster pulls this skill from here and broadcasts it to the AgentCulture mesh — it is NOT vendored from guildmaster like the inbound skills here.","https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes\u002Fblob\u002Fmain\u002F.qwen\u002Fskills\u002Fsummarize-delivery\u002FSKILL.md",[],[],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M7AB5NA0QY9NN076N24NJPJQ","spec-to-plan","Turn a converged devague spec into a buildable plan by working forwards (the spec→plan leg; drives the `devague plan` CLI group). Seed a plan from a converged frame, add tasks that collectively cover every coverage target (the frame's confirmed claims + honesty conditions), give each task acceptance criteria and an honest dependency order, park genuine unknowns as first-class risks, and export a plan only once it *converges*. Use when the user says \"spec to plan\", \"stp\", \"turn this spec into a plan\", \"plan this spec\", \"make a build plan\", or after the `think` skill exports a spec. Authored and maintained in agentculture\u002Fdevague (origin = devague); guildmaster pulls this skill from here and broadcasts it to the AgentCulture mesh — it is NOT vendored from guildmaster like the other skills here.","https:\u002F\u002Fgithub.com\u002Fagentculture\u002Fculture-nodes\u002Fblob\u002Fmain\u002F.qwen\u002Fskills\u002Fspec-to-plan\u002FSKILL.md",[],[]]