knowledge-base-integration
Use when configuring or running product-level Knowledge Ingest, target-ref Feature catalogs, Evidence Policy decisions, Revision Sets, idempotent smoke, or local/remote knowledge runners across repositories.
- License 许可
- Apache-2.0
- 最近更新
- 2026-11-05
- Stars
- ★ 195
- Forks
- 1
01 前言 Front Matter
- name
- knowledge-base-integration
- description
- Use when configuring or running product-level Knowledge Ingest, target-ref Feature catalogs, Evidence Policy decisions, Revision Sets, idempotent smoke, or local/remote knowledge runners across repositories.
02 资源文件 17
- 许可证LICENSE11.1 KB
- 其他NOTICE720 B
- 参考文档references/artifact-manifest.schema.json1.3 KB
- 参考文档references/deployment-manifest.example.yaml658 B
- 参考文档references/deployment-manifest.schema.json1.6 KB
- 参考文档references/evidence-decision.schema.json2.3 KB
- 参考文档references/metrics.schema.json861 B
- 参考文档references/product.example.yaml1.9 KB
- 参考文档references/product.schema.json4.8 KB
- 参考文档references/revision-set.schema.json1.2 KB
- 参考文档references/runner-job.schema.json1.3 KB
- 参考文档references/runner-result.schema.json1.7 KB
- 参考文档references/runtime-contract.md3.8 KB
- 参考文档references/workspace.local.example.yaml790 B
- 参考文档references/workspace.schema.json2.4 KB
- 参考文档requirements.txt36 B
- 脚本scripts/knowledge_base_p1.py90.3 KB
03 SKILL.md 正文
Knowledge Base Integration
Use this Skill for the P1 read-only integration runtime. Repository-owned .feature files at configured target refs remain the behavior SOT. The runtime produces rebuildable catalogs and evidence bundles; it does not change source repositories.
Steps
- Read the product registry and server workspace mapping. Validate both with
scripts/knowledge_base_p1.py validate-config. Completion: every required repository has an explicit remote, role, target ref, feature root, and contained local mapping. - Resolve Evidence intent before a formal run with
decision. Completion: the immutable decision recordsrequired/not-required/blocked, targets, source rule, reason, policy version, and digest. - For
read / 读取, runingest. Completion: every available ref is fixed to a full commit SHA; complete no-ID Gherkin structures, locators, bindings, conflict candidates,metrics.json, andingest-summary.jsonexist;mutationisnone. - For CI or knowledge-server smoke, run
smokeonly after runner, environment, account, data, service, device, and app-artifact prerequisites are known. Completion: staged native commands ran through the local or configured command Runner Adapter; current-run reports, Chinese summaries, artifact manifest, checksums, metrics and evidence bundle exist, or the run is explicitlyblocked. - Report exact status and paths. P1 publication is always
not-configured; PR checks, remote publication, invalidation, retention, quarantine and gates belong to P2.
Commands
python scripts/knowledge_base_p1.py validate-config \
--product /path/to/product.yaml \
--workspace /path/to/workspace.local.yaml
python scripts/knowledge_base_p1.py decision \
--product /path/to/product.yaml \
--repository smart-fuzi-web \
--trigger pull-request \
--execution-profile developer-local
# CI-generated PR evidence resolves the same policy with an explicit CI source.
python scripts/knowledge_base_p1.py decision \
--product /path/to/product.yaml \
--repository smart-fuzi-web \
--trigger pull-request \
--execution-profile ci
python scripts/knowledge_base_p1.py ingest \
--product /path/to/product.yaml \
--workspace /path/to/workspace.local.yaml \
--output /path/to/generated/products/smart
python scripts/knowledge_base_p1.py smoke \
--product /path/to/product.yaml \
--workspace /path/to/workspace.local.yaml \
--output /path/to/runtime/runs/smart-smoke \
--trigger schedule \
--execution-profile knowledge-server \
--suite-key smoke \
--attempt 1
Read runtime-contract.md before configuring retry/attempt identity, staged smoke, report integrity, trusted deployment metadata, or a non-local Runner Adapter. Start deployment metadata from deployment-manifest.example.yaml.
Pass trusted deployment metadata with --deployment-manifest. verified requires its path, issuer and digest to be trusted, its repository SHAs to equal the Revision Set, and its expected runner/tool/image attestation to match the Adapter result. For local execution, workspace.local_runner plus repeatable --runner-label describes current capabilities. For Mobile or another device pool, declare runner and required_runner_labels on the command and configure that runner key in the server Workspace.
smoke --execution-profile accepts ci or knowledge-server; developer-local reports remain owned by project-native validation. A CI smoke envelope uses clean detached worktrees, exact revisions, and Evidence Publication: not-configured until a P2 publisher accepts it.
Use --no-fetch only when the configured refs are already present locally and the caller explicitly wants an offline run. A missing ref is blocked; the runtime never falls back to the current or default branch.
Configuration and Schemas
- Start from product.example.yaml and workspace.local.example.yaml.
- Validate persisted artifacts against the JSON Schemas in
references/, including product/workspace, decision, Revision Set, deployment metadata, Runner job/result, artifact manifest and metrics contracts. - YAML input requires PyYAML; JSON configuration remains available without it. Install the pinned minimums in requirements.txt in the knowledge-server runtime environment.
Hard Boundaries
sync / 同步remains the writable BDD Sync Mode ingherkin-bdd. Invoke this Skill'sread / 读取branch only for explicit read-only intent with no add/change/update/delete intent; mixed read-and-mutate requests remain in the normal BDD workflow.- Existing Feature and Scenario text is parsed without adding IDs, owner tags, scope tags, or a Gherkin runner.
- Long-lived clones are fetch caches. Smoke commands run in detached isolated worktrees.
- Smoke command strings are shell-free argv strings. Use
stagefor preflight, preparation, tests and cleanup; cleanup remains runnable after an earlier stage blocks. - Formal report declarations provide an exact relative path or glob for a runner report and its same-stem Chinese Markdown summary. Only pairs created or refreshed by the current command are collected; stale, non-Chinese, missing, unsafe or sensitive artifacts make Evidence
blocked. - Current report-only smoke continues to emit
schemaVersion: 1withfeatureSources: []. That envelope remains valid generic evidence and must not be treated as BDD scenario coverage. Scenario-backed knowledge evidence requires a separate v2 envelope; do not rewrite v1 smoke into v2 links. - Duplicate logical runs reuse their idempotency record. Use a new
--attemptfor an intentional rerun; infrastructure retries stay inside that attempt and assertion failures do not retry automatically. smoke-only, contract, mock, app-mocked, and backend-only modes retain their actual meaning.- Configuration and artifacts contain no credentials, tokens, PII, production data, or sensitive request content.
Output
Report:
Knowledge Ingest:run/partial/blocked/not-neededEvidence Contract:not-required/required/blockedEvidence IntentandEvidence TargetsRevision Set: id, status, and each repository ref/SHAKnowledge Server Smoke:passed/failed/blocked/not-neededEvidence Source:ci/knowledge-server/not-neededfor smoke runsEnvironment Alignment:verified/unverified/mismatch/not-neededEvidence Publication:not-configuredfor P1- generated catalog, summary, report, and envelope paths
- idempotency key, parser version, suite key and attempt
- artifact manifest, checksums, runner attestation and metrics paths
04 同仓库更多 Skills
gherkin-bdd
Use when adding, changing, reviewing, testing, synchronizing, or read-only ingesting user-visible behavior with BDD, Gherkin, Given/When/Then scenarios, .feature files, acceptance criteria, knowledge catalogs, or bug-fix behavior specs.
sbtd-workflow-onboard
Checks, installs, or resets SBTD workflow tools, global Skills, managed AGENTS templates, and per-project configuration for local project roots in Oh My Pi.
web-ui-autotest-generator
从前端和后端源码中生成可维护的 Web UI 自动化测试。适用于分析页面、路由、组件、接口、用户流程,生成 Playwright UI 测试,运行测试,生成中文报告,并自检页面功能和跨页面逻辑是否被独立测试资产覆盖。 Generate maintainable Web UI automated tests from frontend and backend source code. Use for analyzing pages, routes, components, APIs, and user flows; generating and running Playwright UI tests; producing Chinese test reports; and checking whether page features and cross-page logic are covered by independent test assets.
seo-geo
SEO & GEO (Generative Engine Optimization) for websites. Analyze keywords, generate schema markup, optimize for AI search engines (ChatGPT, Perplexity, Gemini, Copilot, Claude) and traditional search (Google, Bing). Use when user wants to improve search visibility, search optimization, search ranking, AI visibility, ChatGPT ranking, Google AI Overview, indexing, JSON-LD, meta tags, or keyword research.
sbtd-task
Use for SBTD task execution and continuation, explicit default/lite/strict mode selection, task records and status changes, cross-session recovery, and task handoffs. Load strict gates only for strict tasks; ordinary questions and read-only work do not create task files.
project-validation
Use after code changes to choose and run validation commands for Node, JavaScript, TypeScript, Python, Go, Dart, Java, Kotlin, C++, Swift, or Objective-C projects. Prefer project-defined commands and report skipped checks and risks.