Live survey304 models · 14 providers · 23,592 auditsAudited implement · SAFE →What changedHow we verifyllms.txt

Agent skill · Audited from source

autoharness

skill · tigerless-labs/autoharnessMIT licenceAudited 12 Nov 2026 · source 44f266c73aea

★ 11,024pushed today
!CAUTIONGrade D · 63/100 · 25 findings · 44f266c
11,024repo stars · whole tigerless-labs/autoharness

Is it safe? With care — CAUTION, grade D (63/100). We read this skill's source at commit 44f266c73aea and observed filesystem and network access and a hard-coded secret (see the findings). The audit raised 25 findings, 0 critical or high. What it does: Use when users /learn, save, remember, or distill lessons.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “autoharness (tigerless-labs/autoharness)”. Retrieved 12 Nov 2026. https://modelsatlas.com/skills/tigerless-labs/autoharness

Trust audit

Audit 0.4.1 · source read 12 Nov 2026

Trust score63/100!CAUTION D
0 critical0 high23 medium10 low0 info

CAUTION means the audit found things worth knowing before you trust this skill. Grade D is shared by 749 of 18,389 audited skills.

What the source does

  • FilesystemDeclared2 observations
  • NetworkDeclared3 observations
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsFoundSee the findings; values are never shown

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS0
L1 Static sourceSecrets, path traversal, supply chain, hidden codeWARN25
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS0
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (25)

MEDIUMNetwork egressnet.beacon_words · src/autoharness/lib/skills_guard.py:16
F-04 Network egressCWE-200 · CWE-319Layer L1
"exfiltration": [
Why
No further detail recorded for this rule.
Fix
—
MEDIUMNetwork egressnet.beacon_words · src/autoharness/lib/skills_guard.py:19
F-04 Network egressCWE-200 · CWE-319Layer L1
r"(exfiltrate|leak|send|post|upload)\b.{0,30}(http|secret|token|api[_-]?key|password|credential|\benv\b)",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMNetwork egressnet.beacon_words · src/autoharness/lib/skills_guard.py:25
F-04 Network egressCWE-200 · CWE-319Layer L1
r"always\s+(exfiltrate|send|leak|forward)",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.aws · tests/test_promoter.py:445
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
intent = {**_create(), "…": "…"}
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.aws · tests/test_promoter.py:449
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
assert "…" in text and "…" not in text
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.aws · tests/test_redact.py:170
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…": "…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.aws · tests/test_redact.py:187
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.aws · tests/test_redact.py:230
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
raw = "\n"…"-----BEGIN RSA PRIVATE KEY-----", "…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.db_uri · tests/test_redact_url_credentials_quadratic.py:9
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
out = redact.redact("…")
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.db_uri · tests/test_redact_vendor_keys.py:61
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.generic · tests/test_redact.py:322
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
secret = "…"
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.github · tests/test_redact.py:311
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
pat = "…"
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.private_key · tests/test_redact.py:12
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…"
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.private_key · tests/test_redact.py:39
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
_BEGIN, _END = "…", "…"
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.private_key · tests/test_redact.py:130
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
raw = "\n"…"-----BEGIN RSA PRIVATE KEY-----", "…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.private_key · tests/test_redact.py:171
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…": "…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.private_key · tests/test_redact.py:188
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.slack · tests/test_redact.py:173
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…": "…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.slack · tests/test_redact.py:190
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.stripe · tests/test_redact.py:13
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…"
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.stripe · tests/test_redact.py:21
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.stripe · tests/test_redact.py:177
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…": "…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.stripe · tests/test_redact.py:192
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…",
Why
No further detail recorded for this rule.
Fix
—
LOWCode injectioncode.eval_exec · tests/test_promoter.py:55
F-02 Code injectionCWE-78 · CWE-94 · CWE-95Layer L1
body = GOOD_BODY + "\nAvoid eval() and exec() in Python scripts.\n"
Why
evaluates text as code
Fix
remove; use a parser or a dispatch table
LOWCode injectioncode.eval_exec · tests/test_skills_guard.py:40
F-02 Code injectionCWE-78 · CWE-94 · CWE-95Layer L1
@pytest.mark.parametrize("reference", ["eval()", "exec()", "eval( )", "exec(\n\t)"])
Why
evaluates text as code
Fix
remove; use a parser or a dispatch table

Audit 0.4.1 · source sha 44f266c73aea · 12 Nov 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publishertigerless-labs GitHub organizationOwner of the repository this skill is read from
  • Source foldertigerless-labs/autoharnessThe exact folder we audited
  • LicenceMITAs the repository states it
  • Commit pinned for this verdict44f266c73aeaA new audit runs when the source changes
  • Repository statusNot archivedPushed today
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Use when users /learn, save, remember, or distill lessons.

---
name: learn
description: "Use when users /learn, save, remember, or distill lessons."
category: general
---
# Learn: distill this session into the skill library

Review the conversation so far and distill what is worth keeping — through the
standard proposal chain, never by writing files.

1. Identify class-level, reusable lessons: corrected approaches, non-trivial
   techniques, durable user preferences. Skip one-off narratives and
   environment-specific failures.
2. Compare first: read the injected skill index; if an existing skill covers
   the topic, prefer a `patch`/`update` over creating a near-duplicate.
3. Reconcile old with new: when a lesson supersedes an earlier rule, search
   the managed skill trees for contradicting statements and stage updates so
   the new version wins everywhere.
4. Stage every change exclusively via the `stage_skill` tool, one intent per
   lesson, with `reason` and a verbatim `evidence` quote from this session.
   The deterministic promoter validates and lands; rejects are reported at
   the next session start.

If nothing meets the bar, say so and stage nothing.

Excerpt, first 1,120 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 44f266c

Download folder ↗
git clone https://github.com/tigerless-labs/autoharness

Commands as the repository documents them: shown, never run by us. Our verdict covers commit 44f266c73aea only.

Works in which agents?

What its own text says, not a test

AgentStatusMore
Claude CodeMentioned in its text, not testedSkills for Claude Code · About Claude Code
“Mentioned” means the skill’s text names the agent. We have not run it there.

Audit history

Every audit, every source change

  1. !CAUTION D 63 / 100

    first audit · source 44f266c73aea

README badge

For maintainers · shows the latest audit

ModelsAtlas: CAUTION DMODELSATLASCAUTION · D
[![ModelsAtlas: CAUTION D](https://modelsatlas.com/skills/tigerless-labs/autoharness/badge.svg)](https://modelsatlas.com/skills/tigerless-labs/autoharness/)

The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.

Add this badge to your README

[![ModelsAtlas: CAUTION D](https://modelsatlas.com/skills/tigerless-labs/autoharness/badge.svg)](https://modelsatlas.com/skills/tigerless-labs/autoharness/)

Questions about autoharness

What does the autoharness skill do?
Building the self-learning layer next-gen agents will ship with, learning skills from real work.
Is autoharness safe to install?
With care. The audit graded it D (63/100) and found 25 things worth knowing before you trust this skill, listed below with the exact line each was found on.
What can autoharness access on my machine?
The audit observed that it reaches the network and reads or writes files. Each of those is consistent with what it says it does. Secrets in the source: found — see the findings.
Which assistants does autoharness work with?
Its documentation mentions claude-code. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (44f266c73aea), read on 2026-11-12. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.