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 skills add asimons81/hardproof --skill learngit clone --depth 1 https://github.com/asimons81/hardproofWrote 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/asimons81/hardproof/learn)<a href="https://agentmods.dev/skills/asimons81/hardproof/learn"><img src="https://agentmods.dev/badge/skills/asimons81/hardproof/learn/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/asimons81/hardproof/learn"><img src="https://agentmods.dev/badge/skills/asimons81/hardproof/learn.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00025 | $0.00288 |
| Opus 5.5 | $0.00010 | $0.00115 |
| Sonnet 5.5 | $0.00005 | $0.00058 |
| Haiku 4.5 | $0.00003 | $0.00029 |
Grade A, and why
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 1mo 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.
What it actually says
Learn from the Run
Purpose
Decide deliberately whether completed work contains reusable project knowledge without silently changing user instructions or skills.
When to use
Use for the LEARN stage selected by the active run context.
Inputs
Read decisions, corrections, review findings, evidence, remaining risks, and the completion report.
Procedure
- Identify lessons that are specific, reusable, and supported by the run.
- Exclude secrets, private output, transient failures, and guesses.
- Record a learning artifact with
hardproof_record, including source-run provenance and intended scope. - If nothing qualifies, record an explicit skip reason rather than inventing a lesson.
- Do not create or edit global knowledge automatically.
Required records
Standard and Critical require a learning artifact or explicit skip reason. Quick may skip with a recorded reason when its profile path requires one.
Exit criteria
The learning decision is durable, privacy-safe, and linked to evidence. Critical also has a human completion approval.
Failure modes
Do not persist credentials, raw private content, unsupported generalizations, or duplicate guidance. Leave proposals reviewable.
Verification
Confirm provenance and redaction, then call hardproof_transition for COMPLETE. Never declare completion only in prose.
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.
- 1mo ago First seen · 35 lines · 25 tokens per session scan A a2e4b4f9a5d5
learn is a skill published in the GitHub repository asimons81/hardproof (6 stars, last pushed 6d ago), licensed Apache-2.0. It adds 25 tokens to every session and 288 once invoked, about $0.0001 per session on Opus 5.5. 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.
Other skills, from other repositories
notebooklm
Full NotebookLM API via notebooklm-py CLI: create notebooks, add sources, generate podcasts, videos, infographics, slides, quizzes, flashcards, mind maps. Triggers on: "notebooklm", "create a podcast", "audio overview", "generate flashcards", "generate infographic", "/notebooklm".
test-harness
Generates pytest test suites with happy path, edge cases, error conditions, fixture scaffolding, mocks, async patterns. Triggers on: "generate tests", "write tests for", "test this function", "create test suite", "pytest for", "unit tests for", "mock strategy for".
repo-tutor
Use when a user supplies a local repository path or remote Git URL and asks to "teach me this repo", "explain this codebase simply", "show how this system works", "help me use this project", "onboard me to this repository", or "show me how to build on it", especially when the result should be a visual or multipage…
literature-review
Systematic literature review workflow: scope, search (arXiv, Semantic Scholar, Google Scholar), screen, extract, synthesize, and identify gaps. Triggers on: "literature review", "survey the literature", "related work", "systematic review", "synthesize the research", "find papers about", "research gap analysis".
repo-readiness-audit
Use when a user asks whether an identified repository is ready for further development, release work, a new feature, handoff, or a new contributor, requiring a disciplined read-only audit before an evidence-backed verdict.
dont-lie-to-me
Use when the user explicitly wants evidence-disciplined answers that separate observed facts, sourced claims, user reports, inference, unknowns, and contradictions before making strong factual or completion claims.