project-knowledge-capture

Capture durable project knowledge after implementation: goals, context, key entrypoints, CodeGraph findings, decisions, verification results, review conclusions, and follow-up notes into docs/knowledge — so the next developer (or the next session) can answer 'why was this done' and 'what to be careful about' in minutes, not days. Use when development completes and project knowledge should be preserved without secrets or process noise. Do NOT use for: chat summaries, raw session transcripts, secrets, customer data, or production data.

Jump to install Review files

Repository: liyown/superpower-enterprise
Last source activity: June 10, 2026 at 03:35
Detected SKILL.md language: English
Stars: 1
Forks: 0

Install options

The review-first prompt is selected by default. You can switch to a direct command or download a local copy.

Install with an AI assistant

Copy this prompt into the AI assistant you're using.

Follow https://skillsmp.com/skill-install/prompt.md to install only the Agent Skill at https://github.com/liyown/superpower-enterprise/tree/main/skills/project-knowledge-capture.

A direct command skips the review prompt. Inspect the source before running it.

npx skills add https://github.com/liyown/superpower-enterprise --skill project-knowledge-capture

The command stays on one line. Scroll horizontally to inspect it before copying.

Prefer a local copy? Download the files currently available to SkillsMP.

Review the source files

Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.

File Explorer

4 files

SKILL.md

Source instructions · Read-only preview

name: project-knowledge-capture
description: Capture durable project knowledge after implementation: goals, context, key entrypoints, CodeGraph findings, decisions, verification results, review conclusions, and follow-up notes into docs/knowledge — so the next developer (or the next session) can answer 'why was this done' and 'what to be careful about' in minutes, not days. Use when development completes and project knowledge should be preserved without secrets or process noise. Do NOT use for: chat summaries, raw session transcripts, secrets, customer data, or production data.
metadata: {"short-description":"Capture durable project knowledge"}

# Project Knowledge Capture Capture stable project knowledge after a development task. This is not a chat transcript and not a changelog. ## Required Loading Always load `prompts/capture.md`. ## Contract - Default target directory: `docs/knowledge/`. - Create or update `docs/knowledge/index.md`. - Create `docs/knowledge/YYYY-MM-DD-<goal-slug>.md` for the task note. - Record durable knowledge only: context, entrypoints, decisions, constraints, tests, review conclusions. - Preserve CodeGraph provenance. If context was gathered by `rg`/file inspection (no CodeGraph), keep the exact line `CodeGraph unavailable; context was gathered by rg/file inspection.` in the CodeGraph Findings section, with the specific `rg` queries and files consulted on the next line. - Do not write secrets, tokens, customer data, production data, sensitive logs, or speculative claims.

View on GitHub

Back to search Back to installation

data-lineage-trace

Use when you need to prove where a piece of data (PII, PHI, payment data, customer identifier) originates, where it flows through the system, where it lands, and who reads it. Triggers on: data lineage, PII flow, GDPR right-to-erasure, data residency, schema change, retention policy, 'where does field X come from', 'who reads PII column Y', 'walk this field end to end', deletion propagation, audit trail, data classification. Do NOT use for: static code review (use java-code-reviewer, react-code-reviewer, go-code-reviewer, python-code-reviewer, or node-code-reviewer for that), or for design-time brainstorming (use obra/superpowers brainstorming). Complementary to goal-driven-development, which orchestrates the spec-to-code flow; data-lineage-trace is the regulated-data counterpart that runs after a change lands.

2026-06-10

goal-driven-development

Use this skill to orchestrate an existing goal or spec through the dev flow. Load when the user asks for end-to-end feature execution, a goal-to-PR pipeline, or any task that crosses spec → implementation → review → knowledge capture. The skill is a thin router: it names the 6 phases (spec intake, blast-radius check, implementation, verification, review gate, knowledge capture) and the skill that does the work for each phase. Do NOT use for: a single review (load the matching `*-code-reviewer` directly), or for a generic author-craft task like TDD / debugging / brainstorming (use obra/superpowers). Complementary to the 17 enterprise-flow skills; for the planned v0.6.1+ skills, the orchestrator names which of them does the work for each phase.

2026-06-10

enterprise-stack

Use when an agent needs the full enterprise-dev-flow skill set: language-specific code review (java/react/go/python/node), spec-to-code orchestration (goal-driven-development), durable post-impl knowledge (project-knowledge-capture), doc drift linting (spec-doc-linter), and (planned) data lineage / CODEOWNERS / incident trace / compliance / rollout / RFC / observability skills. Triggers on: 'set up enterprise skills', 'install the full stack', 'enterprise-stack', 'give me the augment', 'what skills should I load', 'superpower-enterprise'. Do NOT use for: a single review (load the matching `*-code-reviewer` directly); a generic TDD/debugging/brainstorming request (use obra/superpowers).

2026-06-10

View all from this repository

project-knowledge-capture

Source facts

Install options

Install with an AI assistant

Review the source files

More from this repository

Install with an AI assistant

More from this repository