stack-currency-check
Verify that the patterns the model will use for a framework and version are current per official docs, and cache it as CURRENT_PATTERNS.md for the project. Use when impact-analysis flags greenfield work in an established framework, a new project uses newer-than-cutoff frameworks, or that file is stale (>30 days). Do not use for incremental work with internal precedent, stack choice (use stack-recommend), or fetching an arbitrary URL (use capture-references).
- License 许可
- MIT
- 最近更新
- 2026-10-10
- Stars
- ★ 10
- Forks
- 0
01 前言 Front Matter
- name
- stack-currency-check
- metadata
- { "tools": "Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch", "category": "research-and-sourcing", "provenance": "first-party", "x-wos-profiles": "core, full", "suggested-model": "claude-sonnet-5-5", "multi-repo-aware": "false", "primary-cursor-mode": "Ask", "context-layers-consumed": "memory, retrieved", "context-layers-produced": "retrieved" }
- description
- Verify that the patterns the model will use for a framework and version are current per official docs, and cache it as CURRENT_PATTERNS.md for the project. Use when impact-analysis flags greenfield work in an established framework, a new project uses newer-than-cutoff frameworks, or that file is stale (>30 days). Do not use for incremental work with internal precedent, stack choice (use stack-recommend), or fetching an arbitrary URL (use capture-references).
03 SKILL.md 正文
Act as a senior/staff engineering technology advisor verifying that the patterns about to be used for a given framework+version match current official recommendations.
Goal:
Verify current patterns for the active project's frameworks before greenfield code is written, and persist the result in project-level CURRENT_PATTERNS.md so subsequent tasks inherit verified guidance instead of defaulting to training-data patterns that may be outdated.
Mandatory context bootstrap (before any output):
- Read these sections in
WORKFLOW_OPERATING_SYSTEM.mdfirst:## LLM execution contract## Editor mode policy(mode definitions only; the tool mapping table is lazy-loaded inwos/editor-mode-mappings.mdand needed only for non-Claude-Code tools)## Global output contract(including Adaptive handoff and Mode selection rule)## Cross-cutting workflow guardrails
- Bootstrap tiers: the light-weight commands (
branch-commit,what-next,where-we-at,slice-closure,compact-task-memory) plus the high-frequencyimplement-approved-sliceandsync-task-state(v3 wave1 item D) read the four sections above with two subsections of## Cross-cutting workflow guardrailsskipped:### External web access (centralized)and### Sequencing heuristics (by phase). Everything else is read at every tier, including### Proposal vs approved persistenceand### Substrate peer ownership (per ADR-0034), since all seven write substrate sections and reason about PROPOSED (state-reconcilestays on the full tier for cross-artifact judgment). The full tier is measured at 12109 tokens, the four always-read sections combined; the two skipped subsections are 1,034 of those (measured 2026-09-28), so the reduced tier is about 11,074. The leaf-reviewer tier (verify-against-rubric, ADR-0226) reads only## Global output contract, measured at 5215 tokens, plus its rubric. - Session bootstrap reuse (skip-if-unchanged; v3 wave1 item D): WHEN this conversation already read the bootstrap sections in an earlier turn still VISIBLE in the context window AND
WORKFLOW_OPERATING_SYSTEM.mdhas not changed since, the command MAY skip the re-read and cite the earlier one, emitting one Command transcript line:Bootstrap: reusing turn <N> read, WOS unchanged. Scoped exception to the context-budget re-fetch rule (wos/context-budget.md, "The re-fetch rule"), because these sections are one large, static, byte-identical read repeated every turn; every other tool result still re-fetches. VISIBLE means the section text itself is still present and quotable now, not merely that a record of the earlier read exists: a harness that clears a tool result while the record survives (ADR-0114) has not satisfied VISIBLE, and self-declared memory after a compaction never qualifies. A stateless-per-turn harness is excluded. The transcript line is mandatory; a silent skip is invalid output. - Resolving the spec and
wos/<topic>.md: canonical repository FIRST. Otherwise use active-adapter docs:~/.claude/workflow-docs(CLAUDE_WORKFLOW_DOCS_DIR),~/.cursor/workflow-docs(WORKFLOW_DOCS_DIR; project.cursor/workflow-docson project installs),~/.codex/workflow-docs(CODEX_WORKFLOW_DOCS_DIR), or<kimi-home>/workflow-docs(KIMI_WORKFLOW_DOCS_DIR). Nonempty overrides win. Kimi home:KIMI_CODE_HOME, else existing~/.kimi-code, else existing~/.kimi, default~/.kimi-code. Use one root for both paths; try other adapters if absent or unknown. Report resolved root, fallback to other adapters, or missing mandatory context in### Command transcript. Installed snapshots refresh and prune retired topics only on sync. - Read additional sections only when relevant to this command's role.
- Align all routing recommendations and next-command suggestions with the current command set.
- Official next-command names only: recommendations, including
Run now, use registered names fromcommands/<name>.mdorcommands/<name>/SKILL.md; installed mode uses the available Fhorja command or skill inventory. No invented aliases.Run now: nonewithMode: N/Aends only if no honest next step exists, defined under### Official command names (routing integrity)(ADR-0126). Required inputs: - active task folder path
SOURCE_OF_TRUTH.md(to identify the active framework versions)PROJECT_CHARTER.md(to confirm stack scope and project type)- list of frameworks to verify (from
IMPACT_ANALYSIS.md ## Currency check requiredwhen triggered byimpact-analysis, or user-provided) - existing
CURRENT_PATTERNS.mdat the project level, if any (to identify what is cached vs needs verification)
Task repository files to create or update:
projects/<client>__<project>/CURRENT_PATTERNS.md(project-level cache; gitignored alongsidePROJECT_CHARTER.mdper ADR-0007)TASK_STATE.mdto record verification results and retain or clear theCurrency check requiredblocker according to those results
External web access:
- This command is in the authorized-command set in the spec
## Cross-cutting workflow guardrails ### External web access (centralized), scoped to verifying current framework patterns and version currency. It MUST funnel every verified source intoREFERENCES.md(capture-references entry format, deduplicated by URL) in addition to theCURRENT_PATTERNS.mdcache, so the fetch is indistinguishable in the audit trail from acapture-referencesrun. - Source priority for verification:
- Official framework docs (e.g. nextjs.org/docs, supabase.com/docs, react.dev, tailwindcss.com/docs, stripe.com/docs)
- Official release notes and changelogs (for version-specific changes)
- Official migration guides (when the pattern has shifted)
- AAA company engineering blogs (Vercel, Stripe, Cloudflare, GitHub, Anthropic) for production patterns
- Do NOT use:
- Stack Overflow answers (often outdated)
- Tutorial sites (often outdated, often from older versions)
- Random blog posts
- The model's own training data when the framework version is newer than the training cutoff
Operating rules:
- Do not implement code.
- Handoff: end with the adaptive
### Handoffblock perWORKFLOW_OPERATING_SYSTEM.md## Global output contract(Mode A compact or Mode B full). - First identify each requested framework's active version from
package.json,SOURCE_OF_TRUTH.md, orPROJECT_CHARTER.md(do not guess). Reuse aCURRENT_PATTERNS.mdentry only when it matches that version, hasStatus: verified, and itsAccessed:date is <=30 days old. An unresolved version cannot qualify for reuse. If every requested framework qualifies, skip fetching and cache writes, cite the entries, and apply the state-update rule below; returnNO_OP_TRACEonly when no state update is needed either. - For each framework without a reusable entry:
- Fetch the official docs for the identified version (use WebFetch on docs.<framework>.com or equivalent official source); if the version is unresolved, mark it unverified per the completeness rule below.
- List the current recommended patterns for the use cases this project needs (auth, data fetching, routing, styling, etc.).
- List the deprecated or replaced patterns to AVOID (e.g.
getSessiondeprecated in favor ofgetUser/getClaims). - Note any breaking changes between adjacent versions that affect this project.
- Cite the source URL and access date for each finding.
- Enumerate and mark every framework (completeness, per ADR-0056 / D-2). The result has exactly one row per framework named in the inputs, each tagged
verifiedorunverified:<reason>. Example reasons: version not pinned inSOURCE_OF_TRUTH.mdorpackage.json, official docs unreachable, framework released after the training cutoff with no docs fetched. A framework you could not verify is reported asunverified:<reason>, never dropped. Silently omitting an unverifiable framework is invalid output: it is the gold-standard-audit gap this command exists to prevent. - When a pattern has changed recently (post-training-cutoff or post a major version bump), explicitly flag it: "Model defaults from training data may be outdated for this framework version."
- Cache results in
CURRENT_PATTERNS.mdso subsequent tasks within the same project consume verified guidance instead of re-fetching. - After evaluating cached and fetched results, update
TASK_STATE.mdonly if its state changes. Remove theCurrency check requiredblocker and route toimplementation-planonly when every requested framework's active version is verified. Otherwise retain or add the blocker listing each unresolved framework and reason; route totargeted-questionsfor missing version facts orcapture-referencesfor missing documentation evidence, then rerunstack-currency-checkbefore planning.
CURRENT_PATTERNS.md format:
# CURRENT PATTERNS
## Project
<client>__<project>
## Verified frameworks
### <framework> v<version>
- Status: verified | unverified:<reason>
- Accessed: YYYY-MM-DD
- Source: <official docs URL>
**Current patterns (use these):**
- <pattern 1>: <one-line description> -- <source URL>
- <pattern 2>: ...
**Deprecated patterns (AVOID):**
- <old pattern>: replaced by <new pattern> in v<version>; reason: <why>
**Breaking changes (vN -> vM):**
- <change>: <impact>
**Notes:**
- <anything else worth caching for this project>
---
### <next framework> v<version>
...
Required output:
- Whether
CURRENT_PATTERNS.mdshould be created or updated (SKIPwhen every requested framework has a reusable entry) - Per-framework verification result with: version, source URL, current patterns, deprecated patterns, breaking changes, and a
verified | unverified:<reason>status. One row per requested framework, none dropped. - Exact content for
CURRENT_PATTERNS.md(full document if create; delta block if update; cite entries if skipped) - Exact
TASK_STATE.mdupdate retaining or clearing theCurrency check requiredblocker per the verification results, orSKIPif state is unchanged - Recommended next step per the state-update rule
- Recommended editor mode
Claim grounding (active epistemic humility)
Claim grounding (active epistemic humility). This block governs what you may assert and how you record it. It is keyed to the substrate section you are writing, not to which command is running, and it is INERT on any output that writes none of the claim-bearing sections below. Full contract and rationale: wos/active-epistemic-humility.md.
-
When this applies. This block fires ONLY while you are writing a claim-bearing substrate section:
TASK_STATE.md ## Current known facts,## Risks to watch,## Observations,## Active files in scope,## Canonical decisions;DECISIONS.md ## Locked decisions;IMPLEMENTATION_PLAN.md ## Current gaps,## Risks and mitigations;IMPACT_ANALYSIS.md;EXTERNAL_RESEARCH.md;REFERENCES.md; or any section whose content is a statement a later command or a human decision will act on. WHEN your output writes none of these, this block imposes nothing: skip it and proceed. This is the D-13 inert clause; a fully-grounded or claim-free output pays nothing. -
The unit is the load-bearing claim. A load-bearing claim is one a downstream command or a human decision consumes. A passing aside is not load-bearing; a statement someone will act on is. Apply the rest of this block per load-bearing claim, not per sentence.
-
Ground it or abstain. Before you assert a load-bearing claim, trace it to the enumerable grounded set: a captured
REFERENCES.mdentry, a file read in this session, command output actually seen, or a passing deterministic gate. A claim supported only by model memory is OUTSIDE the grounded set, including when you are right, because that support is not observable. WHEN a load-bearing claim falls outside the set, do NOT assert it: either investigate until it is grounded, or abstain per rule 6. -
Status records provenance, never confidence. WHERE you attach an epistemic status to a claim, the status names WHERE THE CLAIM CAME FROM: a
REFERENCES.mdentry title, a file path plus line, or the gate output it came from. It SHALL NOT express a degree of certainty. Do NOT add a confidence field, a numeric threshold, or a self-assessment prompt anywhere; a self-reported confidence signal is not a usable control signal (wos/active-epistemic-humility.mdPart 1.3). A status whose referent slot is empty is read as UNKNOWN, not as a weak yes. -
Persisted claims carry the status; chat-only claims carry it when they route. Every load-bearing claim you write into a task-memory artifact carries its provenance referent, and that referent travels with the claim so a later command reads it too; do not drop it at the write boundary. A load-bearing claim that appears only in a chat-turn output carries a status only when it crosses the grounding boundary and triggers a route (an abstention, an escalation).
-
Abstain as a routed continuation, never a bare refusal. WHEN you abstain, name the specific investigation that would settle the question AND route to the command that runs it (
capture-references,code-locate,incident-triage, or the fitting one). A withholding that stalls the work is invalid output. Abstention is distinct fromNO_OP:NO_OPmeans there is no work to do; abstention means there is work and the grounding to do it is missing. -
An unfired gate is not evidence. The absence of a fired check does not mean grounding existed. Do not read silence here as a pass.
Standard output layout (required)
Produce the command output using this structure (English only):
Artifact changes
Follow ## Global output contract in WORKFLOW_OPERATING_SYSTEM.md for APPLIED / PROPOSED / SKIP rules. Every listed file carries one of those three tokens, in Lean output too; a prose verb like "written" is not a label.
Command transcript
Brief audit trail (max 4 lines; max 3 in no-op runs with NO_OP_TRACE).
Handoff
Use the adaptive ending format of WORKFLOW_OPERATING_SYSTEM.md ## Global output contract. Every Handoff is one fenced text block with all four lines, Run now:, Mode:, Work complexity: and Reason:, on a stop and on a refusal too; the terminal form is Run now: none with Mode: N/A. A new Mode:, a model or a fresh session is never a stop, an offered choice is one, and Reason: names a role, never a model.
Definition of done (command output)
- Each verified framework has: version, source URL, current patterns list, deprecated patterns list. Vague "best practices" without source URLs is invalid output.
- Every framework named in the inputs appears in the result tagged
verifiedorunverified:<reason>; a requested framework absent from the result is invalid output (the silent-omission gap, per ADR-0056 / D-2). - All source URLs are official docs, official release notes, official migration guides, or AAA company engineering blogs. Stack Overflow links, random blogs, or tutorial sites are invalid.
CURRENT_PATTERNS.mdisAPPLIED.- Cached entries are reused only for matching active versions with
Status: verifiedand anAccessed:date <=30 days old. A full cache hit emitsNO_OP_TRACEonly ifTASK_STATE.mdalso needs no update; otherwise it skips the cache write and reports the state update. - Output ends with a complete
### Handoffblock per the adaptive format. - Before declaring this output done, confirm it satisfies the shared Definition of done (command outputs) and Gate conditions in WORKFLOW_OPERATING_SYSTEM.md.
Quality bar: Optimize for evidence-grounded verification, project-level caching, and preventing the "gold-standard audit" anti-pattern (shipping outdated patterns that need follow-up fix tasks).
04 同仓库更多 Skills
fixture-status
Fixture skill that reads status only. Use when you want a quick status. Do not use when several plans run at once (use fixture-plan-fleet).
fixture-plan
Fixture skill that plans a thing end to end. Use when the plan is missing. Do not use when you only need a status read.
fixture-plan-fleet
Fixture skill, the fleet variant of fixture-plan. Use when several plans run at once. Do not use for a single plan (use fixture-plan).
fixture-status
Fixture skill that reads status only. Use when you want a quick status. Do not use when several plans run at once (use fixture-plan-fleet).
fixture-plan
Fixture skill that plans a thing end to end. Use when the plan is missing. Do not use when you only need a status read (use fixture-status).
fixture-plan-fleet
Fixture skill, the fleet variant of fixture-plan. Use when several plans run at once. Do not use for a single plan (use fixture-plan).