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.
npx agentmods add skills/pratiyush/llm-wiki/self-learnnpx skills add Pratiyush/llm-wiki --skill self-learngit clone --depth 1 https://github.com/Pratiyush/llm-wikiWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/pratiyush/llm-wiki/self-learn)<a href="https://agentmods.dev/skills/pratiyush/llm-wiki/self-learn"><img src="https://agentmods.dev/badge/skills/pratiyush/llm-wiki/self-learn.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00084 | $0.01572 |
| Opus 5 | $0.00042 | $0.00786 |
| Sonnet 5 | $0.00017 | $0.00314 |
| Haiku 4.5 | $0.00008 | $0.00157 |
Grade A, and why
self-learn 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 6d 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.
How it starts
The opening of the file, as written. The whole thing — 131 lines — stays where its author put it; the contents beside it link to each section on GitHub.
self-learn
What this skill does
Closes the Dogfooding Meta-Loop from the Open Source Framework v4.1.
Every non-trivial session on a framework-driven project produces lessons: patterns that worked, patterns that failed, gotchas hit, decisions made. Most tools let those lessons evaporate. This skill captures them, runs a quality gate, and proposes framework updates.
It is the reason the framework evolves. Framework v4.0 → v4.1 was a self-learn pass that folded llmwiki's learnings back into the parent Open Source Framework.
When to invoke
- User says "learn from this session", "what did we learn", "extract lessons", "distill this"
- After a substantial feature ships (especially if it required multiple debugging loops)
- At the end of a project phase — before moving to the next phase
- When the user fixes a bug that was caused by a missing framework rule
- When the user finds a pattern that should apply across projects
- At the end of a "monthly verification" pass (Phase 8 Maintain)
Do NOT invoke when:
- The session is trivial (single file change, typo fix)
- The session is still in progress (wait until the user says "done")
- The learnings are obvious or already codified in the framework
Workflow
-
Gather context. Read:
- The recent session transcript (
raw/sessions/<project>/<latest>.mdor Obsidian session notes) _progress.mdto know the current phasetasks.mdto see what shippeddocs/framework.mdto know the current framework stateCHANGELOG.mdfor what's already been logged
- The recent session transcript (
-
Extract candidate lessons. Look for:
- Failed attempts — "X didn't work because Y" → candidate rule
- Surprising wins — "X worked and I wouldn't have guessed" → candidate pattern
- Repeated debugging loops — "I hit X three times this week" → candidate hard rule
- Decisions — "I chose X over Y because Z" → candidate Project Type addition
- Hints the user gave — "we need to always do X" → candidate steering rule
- Gaps in the roadmap — items discovered during execution that weren't in the plan
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.
- 6d ago First seen · 131 lines · 84 tokens per session scan A 084214c0f682
self-learn is a skill published in the GitHub repository Pratiyush/llm-wiki (383 stars, last pushed 2mo ago), licensed MIT. It adds 84 tokens to every session and 1,572 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
manuscript-reframe
Reframe report-like academic drafts into paper-form scientific arguments while preserving or explicitly renegotiating author intent; requires an approved old-versus-proposed spine, evidence and argument baselines, analysis-role control, and post-edit drift review.
audit
Use when checking a thesis draft before submission for inconsistent numbers, terminology, cross-references, or citation problems.
evidence-review
Build evidence-controlled literature reviews and gap maps with source-status labels, claim registers, citation-role plans, traceability tables, and overclaim audits. Use when drafting or auditing review papers, thesis literature reviews, scoping reviews, or evidence syntheses where adjacent-domain evidence, candidate…
peer-review
Review another author's manuscript, paper, thesis chapter, proposal, or preprint as an external reviewer. Use when asked to evaluate novelty, significance, gap-contribution fit, claim-evidence adequacy, methods, evaluation, overclaim risks, structure, writing, required revisions, or recommendation without rewriting…
revision-escalation
Stop repeated failed writing, coding, manuscript, rebuttal, or restructuring revisions when the same issue has gone through 3+ unsatisfactory edits, vague feedback such as still wrong/weird/unclear/weak/越改越乱, version contamination, or possible gap/claim/evidence/venue-fit drift.
verify
Fact-check claims encountered during reading — dates, names, events, citations. Use when encountering historical facts or disputed claims.