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/everywan-dev/claude-code-engineering/set-up-verified-documentationnpx skills add everywan-dev/claude-code-engineering --skill set-up-verified-documentationgit clone --depth 1 https://github.com/everywan-dev/claude-code-engineeringWhat 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 | $0.00087 | $0.00661 |
| Opus 5 | $0.00044 | $0.00331 |
| Sonnet 5 | $0.00017 | $0.00132 |
| Haiku 4.5 | $0.00009 | $0.00066 |
Grade A, and why
set-up-verified-documentation 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.
How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Set up verified documentation
Adopting the plugin in a project is one command, run from the project root.
Bootstrap the layout
python3 -m validated_memory init
This scaffolds, creating each item only if missing (existing items are never touched):
knowledge/-- empty; curated-knowledge units go here.memory/-- empty except for its index,memory/MEMORY.md.validated-memory.md-- the adopter configuration: declares the extension, theid_prefix, and the probe registry (already mappinggit_refto the bundled probe).knowledge-extension.md-- a valid, empty declared extension (fields: []).
init reports init: created <path> or init: kept <path> for each item,
and is safe to re-run: it is idempotent and never overwrites a hand-edited
file. See the README's init section for the full contract, including
--harness-memory below.
Wire the harness's persistent memory (optional)
python3 -m validated_memory init --harness-memory PATH
Makes PATH a move-proof symlink to this project's memory/ directory, so
the harness reads agent memory from wherever it expects it while the data
stays versioned inside this repo. Safe to call repeatedly, including after
the project is renamed or re-cloned -- it only ever re-points the symlink,
never deletes data. The plugin's SessionStart hook
(hooks/restore-memory-symlink.sh) already calls this automatically for an
adopted project on every session start; running it by hand is only needed to
wire a harness location the hook does not already know about.
Verify the adoption
Right after init, both enforcement commands must pass clean:
python3 -m validated_memory validate
python3 -m validated_memory lint
validate may still report a WARNING for an empty knowledge/ (no units to
check) -- that does not gate. Any ERROR means the scaffold is broken; do not
proceed until both commands are clean.
Next steps
- Declare adopter-specific fields by editing
knowledge-extension.md-- see thedocument-with-evidenceskill and the README's "Declared extension" section. - Register a probe for each anchor
kindyour units will use, by adding an entry underprobes:invalidated-memory.md-- see thedetect-stale-documentationskill and the README'sprobesection. - Start writing curated knowledge with the
document-with-evidenceskill.
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.
- 2d ago First seen · 68 lines · 87 tokens per session scan A 7328a1b86461
set-up-verified-documentation is a skill published in the GitHub repository everywan-dev/claude-code-engineering (2 stars, last pushed 13d ago), licensed Apache-2.0. It adds 87 tokens to every session and 661 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 2026-08-31.
Other skills, from other repositories
skill-builder
Automatically detect source types and build AI skills using Skill Seekers. Use when the user wants to create skills from documentation, repos, PDFs, videos, or other knowledge sources.
sandbox-environment
Sandbox API reference, proven patterns, and library recommendations for evaluateexpression.
generate-ai-rules
Generate AI assistant configuration files for a repository — CLAUDE.md, AGENTS.md, and Cursor rules (.cursor/rules/.mdc) — from codebase analysis. Use whenever the user wants to create or update CLAUDE.md, AGENTS.md, agent rules, Cursor rules, AI coding assistant configuration, or "onboard AI tools" to a project, even…
learn-from-correction
Propose a principle edit to a skill or persona file based on a (agentoutput, humanoutput) correction pair. Outputs a proposal markdown for human review - never auto-edits the target file.
ai-rulez
Manage AI assistant governance rules across Claude, Cursor, Windsurf, Copilot, Gemini, and other tools using ai-rulez. Use when configuring rules, context, skills, domains, profiles, includes, plugins, or generating tool-specific outputs.
mstar-skill-authoring
Agent skill 撰写 / 重写 / 优化规范（SkillsBench 实验门控）。在新建任意 skill、大改 SKILL.md、改写 description 触发契约、压缩过长 body、评审 skill 结构、或判断某段行为塑形文案是否值得保留时读取。适用于任何领域的 skill，不限于本仓库专题；不用于普通应用实现。.