knowledge-capture

A tool for saving important project knowledge from a work session as scoped rule files. The files are placed under .claude/rules/ so they can guide later work in relevant parts of the project.

In plain words
What is it for?
Use it after discovering reusable project behavior, framework quirks, data flows, recovery procedures, or other details that would help future coding sessions.
Why use it?
It prevents valuable discoveries, recurring workarounds, and complex system details from being forgotten after a session. It also filters out one-off fixes and generic knowledge that do not need to be saved.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/aigentive/ralphx/knowledge-capture
Any agent
npx skills add aigentive/RalphX --skill knowledge-capture
Clone the repo
git clone --depth 1 https://github.com/aigentive/RalphX

Made for: Claude Code, Codex.

Per session 70 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 805 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00070 $0.00805
Opus 5 $0.00035 $0.00402
Sonnet 5 $0.00014 $0.00161
Haiku 4.5 $0.00007 $0.00081

Measured 2d ago against content hash 1878b16decda, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

knowledge-capture scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/app/skills/knowledge-capture/SKILL.md · 109 lines

How it starts

The opening of the file, as written. The whole thing — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Knowledge Capture

Packages session learnings into scoped .claude/rules/ files.

Step 1: Read Framework Reference

Read the memory framework doc for knowledge capture criteria and format: ${CLAUDE_PLUGIN_ROOT}/memory-framework.md

Also fetch/refresh the official Claude Code memory docs (cached locally):

"${CLAUDE_PLUGIN_ROOT}/skills/rule-manager/scripts/rule-fetch-docs.sh"

Step 2: Evaluate Session

Review the current session's conversation to identify:

Worth Capturing

  • Complex multi-component interactions (e.g., "Component A triggers Event B which updates State C")
  • Error recovery procedures that took multiple attempts to resolve
  • Deep internal system knowledge (state machines, data flows, edge cases)
  • Framework/library quirks and workarounds (e.g., React ref equality, Tauri serialization)
  • Patterns that would save >15 minutes if known upfront next time

NOT Worth Capturing

  • One-off fixes unlikely to recur
  • Knowledge already documented in existing .claude/rules/ files
  • Generic programming knowledge (things Claude already knows)
  • Temporary workarounds that should be fixed properly
  • Simple bug fixes with obvious causes

If nothing worth capturing -> output "No significant learnings to capture." and stop.

Step 3: Classify Each Learning

For each learning worth capturing, determine:

3a. Scope (which files does it relate to?)

  • Identify the files that were involved in the discovery
  • Determine glob patterns that cover those files
  • Example: discovery about TaskGraph -> paths: ["src/components/TaskGraph/**"]

3b. Existing Rule Check

Run the audit script to see current rules:

"${CLAUDE_PLUGIN_ROOT}/skills/rule-manager/scripts/rule-audit.sh" --json

Check if any existing rule file covers the same domain:

  • If yes -> append to that file (add new ## section)
  • If no -> create new file

3c. Format Entry

Use the auto-memory style format:

## [Descriptive Title]
- **Problem**: What went wrong or was complex
- **Fix**: How it was resolved
- **File**: Key file paths involved (e.g., `src/components/TaskGraph/TaskGraphView.tsx`)
- **Pattern**: Reusable principle (if applicable)

Read the full file on GitHub · 109 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 109 lines · 70 tokens per session scan A 1878b16decda

Subscribe to this mod's changes

knowledge-capture is a skill published in the GitHub repository aigentive/RalphX (5 stars, last pushed 2d ago), licensed Apache-2.0. It adds 70 tokens to every session and 805 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens