Forensify
alexgreensh/repo-forensics
Cross-agent self-inspection of your AI-agent stack. An agent skill from alexgreensh/repo-forensics.
Hunt for the failure this project exists to prevent — code that compiles, types, tests green, and quietly reports "all clear" about something it never examined.
Install False Green Audit
$ npx skills add guardana/guardana --skill false-green-audit -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install guardana/guardana false-green-audit --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/guardana/guardana.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/false-green-audit .claude/skills/false-green-audit && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "false-green-audit" agent skill from https://github.com/guardana/guardana/tree/main/.claude/skills/false-green-audit into .claude/skills/false-green-audit/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "false-green-audit", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/guardana/guardana/tree/main/.claude/skills/false-green-auditType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add guardana/guardana --skill false-green-audit -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install guardana/guardana false-green-audit --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/guardana/guardana.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.claude/skills/false-green-audit .agents/skills/false-green-audit && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "false-green-audit" agent skill from https://github.com/guardana/guardana/tree/main/.claude/skills/false-green-audit into .agents/skills/false-green-audit/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "false-green-audit", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add guardana/guardana --skill false-green-audit -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install guardana/guardana false-green-audit --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/guardana/guardana.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.claude/skills/false-green-audit .cursor/skills/false-green-audit && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "false-green-audit" agent skill from https://github.com/guardana/guardana/tree/main/.claude/skills/false-green-audit into .cursor/skills/false-green-audit/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "false-green-audit", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/guardana/guardana.git --path .claude/skills/false-green-audit--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add guardana/guardana --skill false-green-audit -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install guardana/guardana false-green-audit --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/guardana/guardana.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.claude/skills/false-green-audit .gemini/skills/false-green-audit && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "false-green-audit" agent skill from https://github.com/guardana/guardana/tree/main/.claude/skills/false-green-audit into .gemini/skills/false-green-audit/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "false-green-audit", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install guardana/guardana false-green-auditInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add guardana/guardana --skill false-green-audit -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/guardana/guardana.git skills-src && mkdir -p .github/skills && cp -r skills-src/.claude/skills/false-green-audit .github/skills/false-green-audit && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "false-green-audit" agent skill from https://github.com/guardana/guardana/tree/main/.claude/skills/false-green-audit into .github/skills/false-green-audit/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "false-green-audit", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add guardana/guardana --skill false-green-audit -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install guardana/guardana false-green-audit --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/guardana/guardana.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.claude/skills/false-green-audit .opencode/skills/false-green-audit && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "false-green-audit" agent skill from https://github.com/guardana/guardana/tree/main/.claude/skills/false-green-audit into .opencode/skills/false-green-audit/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "false-green-audit", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
false-green-audit- GitHub stars
- 259
- Token cost
- ~1.1k tokens
- SKILL.md length
- 704 words
- Files
- 1
- Skills in repo
- 13
- Repo updated
- First seen
- Licence
- Apache-2.0
At a glance
Hunt for the failure this project exists to prevent — code that compiles, types, tests green, and quietly reports "all clear" about something it never examined.
- Reviewing a release
- SKILL.md covers The seven shapes, with the…, The method that has found the…, Where to look first and When you find one
- Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
- Auditing a subsystem
What it does
False Green Audit is an agent skill from guardana/guardana. Hunt for the failure this project exists to prevent — code that compiles, types, tests green, and quietly reports "all clear" about something it never examined. Use when reviewing a release, auditing a subsystem, or before tagging.
Its SKILL.md is about 1.1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It sits in Security, covering Prompt injection and agent security. It works with Model Context Protocol. The repository describes itself as: Open-source AI security verification for model artifacts, live endpoints, MCP servers, and recorded agent traces. Reproducible evidence for release decisions. The licence is Apache-2.0.
When your agent uses it
- Reviewing a release
- Auditing a subsystem
Example prompts
- “all clear”
- “/false-green-audit”
What it can do on your machine
Read from SKILL.md and the folder at commit e76cd24. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
No scripts in the folder and no shell commands in SKILL.md.
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
False Green Audit loads about 1.1k tokens when it runs. Until then it costs about 62 tokens; SKILL.md has 704 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from guardana/guardana at commit e76cd24, republished under its Apache-2.0 licence (© guardana). 704 words, ~1,137 tokens.
Download SKILL.mdSave it as
.claude/skills/false-green-audit/SKILL.md (or your agent's skills folder).- name
- false-green-audit
- description
- Hunt for the failure this project exists to prevent — code that compiles, types, tests green, and quietly reports "all clear" about something it never examined. Use when reviewing a release, auditing a subsystem, or before tagging.
Hunting a false green
Every audit of this repository has found real defects on top of a fully green gate. That is not a comment on the gates; it is what this class of bug is. The code is correct in every way a linter, a type checker or a unit test can see, and wrong in the one way that matters: it says "nothing found" about something it did not look at.
Treat a green gate as the start of an audit, never its conclusion.
The seven shapes, with the release each was found in
1. Silence spelled pass. A check that cannot actually run returns clean
instead of inconclusive. Grep for except ... : pass, for return () in a
branch that means "I could not tell", and for any evaluator path that yields a
verdict without having seen text. An empty reply graded pass@0.95 (0.12).
2. A channel rebuilt field by field. A function that constructs a dataclass by
listing its fields silently drops the next field somebody adds. Look for
return SomeResult(a=..., b=..., c=...) where the input was already a
SomeResult; it should be replace(...). compare_reports dropped the whole
measurement channel (0.22); ScanResult.merged exists because errors went
missing the same way (0.9).
3. A whitelist where a scan belongs. A gate that iterates a hand-written list
of files covers the files somebody remembered. Every new file is invisible.
Image pins sat on :0.9 for twelve releases because the four files carrying them
were created after the list (0.22).
4. A promise that rots. A statement that was true when written and became false with no diff to blame: "coming in v0.7", "the collector has no audit log", a pin to a container tag. Five future-tense claims about features shipped fourteen releases earlier (0.22).
5. A seam nothing exercises. A documented extension point with no registrant
anywhere is a seam nobody has run. guardana.targets was in the contract from
0.1 with no example, so Registry.targets() returned [] and a false red
shipped in pack validate (0.18); guardana.taxonomies was in the same state
until 0.19.
6. A test that measures an echo. An assertion on a document, a log line or a mock's call count measures what the code said, not what it did. Move the assertion to the seam where the value has to arrive. A capability check that asserted on the manifest instead of on what ran.
7. A test that cannot fail. Invert the behaviour, not the branch: change
the production code so the thing under test is genuinely wrong, and confirm the
test goes red. A test built on getattr(x, "thing", ()) where nothing has
thing is vacuous and looks thorough.
Show full SKILL.md (262 more words)Show less
The method that has found the most
Run the documented command and read the artifact. Not the test suite — the command a user would type, against a real (or realistically faked) target, and then open the JSON it wrote. Three separate releases had their worst defect found this way and by nothing else:
probe --max-requests 5sent ten (0.12);- a 66 KB
model.pthidposix.systembehind 65 MB of deflated padding and produced zero findings (0.12); diffreported "no regression" over anindeterminaterun (0.17);compare_reportsdropped the measurement channel — every unit test passed because each tested the half it owned (0.22).
A fake endpoint is three lines of http.server; there is no excuse for skipping
this step.
Where to look first
The seam between what the project claims and what it does: documentation against behaviour, a schema against its reader, a capability against its surface, a manifest field against whatever is supposed to populate it. Both of the two worst findings in the 0.21 audit were there, and neither was in the engine.
Then: anything that reads a file somebody else wrote, anything with a version
or schema_version, anything that decides pass vs indeterminate, and any
field of a persisted document that no fixture populates.
When you find one
Fix it and add the gate that makes it impossible — a whitelist becomes a scan, a promise becomes a test, a hand-written list becomes a measurement. The repo's own rule: a finding is closed when it has a gate that will not let it back, not when it has been corrected once.
© guardana, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
Just SKILL.md in .claude/skills/false-green-audit of guardana/guardana.
Open the folder on GitHubat commit e76cd24
Compare with similar skills
False Green Audit next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| False Green Audit this skillguardana/guardana | 259 | — | ~1.1k | Automated safety check: Pass | Apache-2.0 | |
| Forensifyalexgreensh/repo-forensics | 190 | — | ~2.5k | Automated safety check: Notes | Custom licence | |
| Hol Guardhashgraph-online/hol-guard | 845 | — | ~542 | Automated safety check: Pass | Apache-2.0 | |
| Setuphashgraph-online/hol-guard | 845 | — | ~443 | Automated safety check: Pass | Apache-2.0 | |
| Statushashgraph-online/hol-guard | 845 | — | ~231 | Automated safety check: Pass | Apache-2.0 | |
| Plugin Scanneriflytek/skillhub | 5.2k | 2 repos | ~1.1k | Automated safety check: Notes | Apache-2.0 |
Similar skills
Hol Guard
hashgraph-online/hol-guard
Run HOL Guard scanner and guard operations via uv run hol-guard.
Setup
hashgraph-online/hol-guard
Install or initialize HOL Guard local runtime protection for Claude Code.
Status
hashgraph-online/hol-guard
Check HOL Guard local protection status for Claude Code without changing configuration.
Plugin Scanner
iflytek/skillhub
Scan AI agent skills, plugins, MCP servers, and agent tooling for prompt injection, unsafe commands, secret exposure, and supply-chain risks before installing or trusting them.
Security Guide
jnMetaCode/shellward
OpenClaw 安全部署指南 / Security deployment guide — help users secure their OpenClaw installation
More from guardana/guardana
All 13 skills in this repoShip
guardana/guardana
Commit and push a finished change the way this repo requires — explicit paths, one commit per logical change, no attribution, the five documentation places answered, the site checks before a push (a…
Add A Rule
guardana/guardana
Add security coverage to Guardana the way this repository requires — as a rule, evaluator or target, never by patching the engine — with the fixtures, the framework mapping and the documentation…
Auto
guardana/guardana
Autonomous, non-interactive run of the whole lifecycle for one development task — size, plan, build, gate, review, fix, commit — without stopping for questions.
Docs
guardana/guardana
Documentation work in this repo — the five places a user-visible change must answer, the page conventions the site build enforces, the tests that pin prose to the registry, and a simplification pass…
Gate
guardana/guardana
Run this project's verification — the full local CI mirror (ruff, mypy, import contract, pytest with PostgreSQL, coverage floors, dogfood, generated docs and site, the isolated example suites, the…
Plan
guardana/guardana
Turn a development task into one work file in .work/ — goal, decisions, lanes with exact files and verification, done-criteria.
Related
Works with
Categories
Questions about False Green Audit
What does False Green Audit do?
Hunt for the failure this project exists to prevent — code that compiles, types, tests green, and quietly reports "all clear" about something it never examined. False Green Audit is an agent skill from guardana/guardana. Hunt for the failure this project exists to prevent — code that compiles, types, tests green, and quietly reports "all clear" about something it never examined.
When should I use False Green Audit?
False Green Audit fits situations like: reviewing a release; auditing a subsystem.
How do I install False Green Audit in Claude Code?
Run `npx skills add guardana/guardana --skill false-green-audit -a claude-code`. Or copy the skill folder (.claude/skills/false-green-audit in guardana/guardana) into .claude/skills/false-green-audit in your project. Claude Code loads it when a task matches its description.
How do I install False Green Audit in Codex?
Run `npx skills add guardana/guardana --skill false-green-audit -a codex`. Or copy the skill folder (.claude/skills/false-green-audit in guardana/guardana) into .agents/skills/false-green-audit in your project. Codex loads it when a task matches its description.
Can I use False Green Audit in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add guardana/guardana --skill false-green-audit -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/false-green-audit, .gemini/skills/false-green-audit, .github/skills/false-green-audit and .opencode/skills/false-green-audit in your project.
What does False Green Audit need to run?
SKILL.md names no scripts, command-line tools or credentials: False Green Audit is instructions for the agent only.
Does False Green Audit access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is False Green Audit safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does False Green Audit use?
False Green Audit is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does False Green Audit use?
About 1.1k tokens (SKILL.md is roughly 4.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to False Green Audit?
Skills that share tags, products or a category with False Green Audit: Forensify (alexgreensh/repo-forensics, 190 stars), Hol Guard (hashgraph-online/hol-guard, 845 stars), Setup (hashgraph-online/hol-guard, 845 stars) and Status (hashgraph-online/hol-guard, 845 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains False Green Audit?
guardana (a GitHub organization) maintains it in guardana/guardana, which has 259 GitHub stars. The repository holds 13 skills in this directory. The repository was last updated on October 11, 2026.
Source: guardana/guardana on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.