菜单
Tier-1 orchestrator for reading the ledger. Stats, reports, flake detection, coverage gaps, historical queries. Never writes — only reads. Use when: "has this passed recently", "flake rate", "show me the last N runs", "coverage gaps", "generate a report", "stats", "exploratory session", "/wicked-testing:insight".
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Evidence-gated acceptance testing with three-agent separation of concerns. Writer designs test plans, Executor collects artifacts, Reviewer evaluates independently. Eliminates false positives from self-grading. Use when: "acceptance test", "verify it works", "did it pass", "run acceptance", "test this scenario", "acceptance criteria", "validate the feature", "/wicked-testing:acceptance"
Record a claim-backing artifact in the vault and attach a deterministic verifier. Use when capturing evidence that "tests pass", "build clean", a commit exists, or a file's contents back a claim — and when replacing stale evidence via supersede. Covers --run vs --artifact, verifier syntax, and contract pinning.
Repo-local structural validator for wicked-testing. Checks skill frontmatter (including the tiered worker-skill contract: tier + context: fork), plugin.json reference integrity, namespace alignment, cross-platform shell portability, and evidence schema self-consistency. Zero LLM cost. Run before publishing or after structural changes. Use when: validating wicked-testing's own skills/plugin.json, pre-publish sanity check, after adding new skills to this repo.
Accessibility specialist — axe-core + pa11y, WCAG 2.1 AA / 2.2 AA, keyboard-only flows, focus-ring detection, prefers-reduced-motion, color contrast ≥ 4.5:1. Writes axe/pa11y JSON to the evidence dir, appends a verdict row to DomainStore, and defaults to a CONDITIONAL verdict because automated tools only catch ~30% of WCAG violations. Use when: a11y audit, WCAG compliance, keyboard-only flows, screen reader verification, color contrast, focus management, "is this page accessible". <example> Context: Reviewer wants a WCAG 2.1 AA check on a new checkout flow. user: "Run an accessibility pass on https://staging.example.com/checkout." <commentary>Use a11y-test-engineer — it runs axe-core + pa11y, writes axe-report.json + pa11y-report.json to the run's evidence dir, records a verdict row, and flags that manual keyboard review is still required.</commentary> </example>
Follows structured wicked-testing test plans step-by-step, collecting evidence artifacts. Executes and captures only — does not judge or grade pass/fail. Writes evidence files to .wicked-testing/evidence/{run-id}/. Use when: acceptance test execution, evidence collection, test plan execution <example> Context: Test plan is ready and needs to be executed step by step. user: "Execute the acceptance test plan for the file upload feature." <commentary>Use acceptance-test-executor for mechanical step execution and evidence capture without judging results.</commentary> </example>
Evaluates evidence artifacts against test plan assertions independently. CRITICAL ISOLATION: Receives ONLY evidence file paths. Never sees execution context. Catches semantic bugs that self-grading misses. Use when: acceptance test review, evidence evaluation, test verdict <example> Context: Executor produced evidence and it needs independent evaluation. user: "Review the evidence from the file upload acceptance tests and render a verdict." <commentary>Use acceptance-test-reviewer for independent, unbiased verdict on test evidence.</commentary> </example>
| name | wicked-testing-insight |
| description | Tier-1 orchestrator for reading the ledger. Stats, reports, flake detection, coverage gaps, historical queries. Never writes — only reads. Use when: "has this passed recently", "flake rate", "show me the last N runs", "coverage gaps", "generate a report", "stats", "exploratory session", "/wicked-testing:insight". |
| argument-hint | [question] [--project <name>] [--json] |
The read-only lens on wicked-testing's ledger. Built on the fixed-SQL oracle so answers are auditable, not LLM-guessed.
/wicked-testing:insight "<question>" [--project <name>] [--json]
Flags --project, --status, --since, and --json are honored as filters
on the underlying oracle queries.
| Input | Dispatch |
|---|---|
| A question, natural language | wicked-testing:test-oracle (fixed SQL) |
| "generate a report" / "run summary" | Oracle rows rendered as a markdown summary (no new SQL); add Tier-2 specialists when the report needs them |
| "find flaky tests" | wicked-testing:flaky-test-hunter |
| "find untested legacy code" | wicked-testing:coverage-archaeologist |
| "run an exploratory session" | wicked-testing:exploratory-tester |
| "audit production quality" / post-deploy read | wicked-testing:production-quality-engineer |
| Unknown question | Oracle returns the supported question list |
| "domain health" / "row counts" / "schema version" | wicked-testing:test-oracle → row_counts / schema_version queries |
| "what tasks are open" / "tasks for X" | wicked-testing:test-oracle → tasks_by_status / tasks_for_project |
Every id in the tables above is a forked worker skill (context: fork) —
invoke it with the Skill tool so it runs in an isolated context:
Skill(
skill="wicked-testing:test-oracle",
args="""Answer the question below against the wicked-testing ledger.
## Question
{natural-language question}
## Optional filters (from flags)
- project: {name or null}
- scenario: {name or null}
- since: {ISO date or null}
## Instructions
1. Route the question to a named query in lib/oracle-queries.mjs by keyword
matching. NEVER synthesize SQL.
2. If no match, return the list of supported question patterns and exit —
do not fabricate results.
3. If better-sqlite3 is unavailable, return ERR_SQLITE_UNAVAILABLE exactly.
4. Run the named query with bound parameters. Return rows as JSON or markdown
table (per --json flag).
5. Include the query name used so the caller can audit.
Do NOT perform state mutations. Do NOT emit bus events."""
)
Swap the skill id to the specialist when the trigger matches something the
oracle doesn't cover — flake detection, coverage archaeology, and exploratory
sessions all have dedicated worker skills.
Insight is heavier on Tier-2 than other Tier-1 skills because most history questions have a specialist answer:
| Trigger | Specialist |
|---|---|
| Flake rate / quarantine proposal | wicked-testing:flaky-test-hunter |
| Coverage gaps, dead-code detection | wicked-testing:coverage-archaeologist |
| Charter-driven exploratory session | wicked-testing:exploratory-tester |
| Post-deploy quality, canary read | wicked-testing:production-quality-engineer |
| Observability assertion audit (trace/log/metric coverage) | wicked-testing:observability-test-engineer |
| Contract drift report (historical / consumer-side) | wicked-testing:contract-testing-engineer |
| "Most impactful tests for HEAD" / TIA lookup | wicked-testing:test-impact-analyzer |
| "Should we ship v2.4.0?" / release readiness | wicked-testing:release-readiness-engineer |
| "Any incident without a regression scenario yet?" | wicked-testing:incident-to-scenario-synthesizer |
The oracle never generates SQL. It keyword-matches the question to one of the
named parameterized queries in lib/oracle-queries.mjs.
If nothing matches, it returns the list of supported questions — it never
guesses. See also lib/domain-store.mjs for the
table-name allowlist that backs the CRUD layer.
This skill does not emit bus events — it is read-only and should not mutate state.
insight is the single read-only door to the ledger. What used to be separate
commands are now questions you ask it (routing per the dispatch table above):
| Old command | Ask insight instead |
|---|---|
stats | "domain health" / "row counts" / "schema version" — routes to the row_counts, schema_version, recent_runs oracle queries |
report | "generate a report for <project/scope>" — oracle rows rendered as a markdown summary |
oracle | any plain-language data question — this is the oracle; insight was always built on it |
tasksis removed, not absorbed.insightis read-only and will not create or mutate tasks. Existing task rows remain queryable here ("what tasks are open?", "tasks for ") via thetasks_by_statusandtasks_for_projectoracle queries. Task creation/update via a command is gone in 0.4.0 — use theDomainStoreAPI (lib/domain-store.mjs) directly if a workflow needs it.
docs/INTEGRATION.mdskills/test-oracle/SKILL.mdlib/oracle-queries.mjs (internal — query catalog)