praneybehl/llm-wiki-plugin / skills/llm-wiki
llm-wiki
VerifiedBuild and maintain an LLM-curated knowledge base from papers, articles, transcripts, notes and project findings. Ingest sources into linked Markdown, answer questions with citations, lint or upgrade a wiki, and preserve useful synthesis across sessions and agents. Also capture verified task experience, consolidate success/failure patterns, and propose, evaluate, apply or roll back evidence-linked skill improvements. Trigger on "add this to my wiki", "what does the wiki say", "learn from this task", "improve this procedure", /wiki:learn or /wiki:evolve. Works on a local folder or a hosted wiki URL (MCP). Use index-first navigation, atomic pages and hybrid search.
assets/evolution_README.md.template
View on GitHub# Skill evolution records
This directory contains canonical local experiment records, not a disposable cache. Back it up with the wiki. `wiki_evolve.py` creates one directory per experiment with complete baseline/candidate skill snapshots, motivating evidence, a diff, evaluation results, and apply/rollback receipts. Never edit snapshots or evaluation results to make a proposal pass.
Search, lint, statistics and graph compilation exclude this directory. Curate useful lessons as ordinary source/concept pages and publish an experiment summary as a synthesis page, linked from the normal wiki index. Do not expose private traces or results in public plugin releases.
Only `apply` changes the selected installed skill after a passing evaluation. A rejected candidate leaves it unchanged. `rollback` restores an accepted change when the skill still matches its snapshot. Both refuse concurrent edits. If a process dies during a transition, rerun the same command to finish it. If a dead process leaves `lock/`, confirm no evolution process is running before removing that empty directory.
See the installed skill's `references/evolution-workflow.md` for capture, consolidation, evaluation, promotion and recovery. An agent runner is a trusted local program. The copies supplied to it are not an operating-system sandbox.
Apply/rollback also use a temporary `.wiki-evolve-lock/` in the target skill to coordinate changes from different wikis. After a process dies, inspect both this lock and the wiki lock before removing stale empty lock directories and retrying. This lock directory is excluded from skill snapshots.
Each started trial also preserves its suite/runner configuration and a complete frozen `corpus/` under the experiment directory. Keep these private alongside the skill snapshots. Corpus hashes are rechecked during promotion. An interrupted trial is reported as incomplete and requires a new proposal ID; do not reuse partially recorded input artifacts.
`wiki_evolve_loop.py` also stores bounded evolution runs here. `learning.json` preserves training patterns and proposal history across runs. Final-test results remain in each run archive and must never be fed back into selection. Raw training observations live in the configured raw root. Run budgets count every role and final/transfer call.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |