SKILLEMALL.ai
GitHub ↗ RU Sign in

AC tracebit-canaries

Use when the user wants to protect their workspace from credential theft, prompt injection, or data exfiltration — even if they don't mention "canaries" or "honeytokens" directly. Covers deploying Tracebit security canary tokens (fake decoy credentials that alert on use), detecting when they're triggered via the user's pre-authorized email tool, and human-supervised incident response. Also use when investigating a suspected compromise, hardening an agent's environment, or setting up tripwires to catch unauthorized access to sensitive files.

ClawHub Agent Skills author: alessandro-brucato v1.0.21 MIT-0 13 files · 4 scripts body ≈ 2 612 tokens Open the sourceclawhub.ai↗ analyzed 3 d ago

As a process C 64/100 · Has gaps — weak spots: result and completion, inputs and preconditions, consistency

AnalyzerInfrastructureAI and agentstype and topics are labelled automatically from the skill text
Runs in: OpenClaw Claude / Claude Code ChatGPT / Codex Gemini CLI Cursor GitHub Copilot / VS Code Qwen Code Kimi Code CLI Z.ai (GLM) DeepSeek (Deep Code) Mistral Vibe LangChain (deepagents) Goose OpenCode
JSON
Technical rating
A
92/100
safety, quality, tests
Safety 60%
94
Quality 40%
90
Run on models
none yet
Process rating
C
64/100
Has gaps
Inputs and preconditions w 11
0
Running it twice w 4
30
Result and completion w 14
40
the three weakest of ten parameters · all ten

How to improve

    For the model run — optional
    • Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
    • A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.

    Guard findings · 6

    ✓ No critical or high findings

    Medium and low: 6
    • low Secrets in code secret-private-key references/api-reference.md:109
      Private key material (placeholder value)
      "sshPrivateKey": "-----BEGIN PRIVATE KEY----- …
      placeholder
    • low Risky intent intent-offensive-security references/canary-types.md:39
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      - Credential harvesting from disk
    • low Risky intent intent-offensive-security references/incident-response-playbook.md:149
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      - Password canary: credential harvesting from files or password manager
    • low Risky intent intent-offensive-security references/incident-response-playbook.md:198
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      | **aws** | AWS credentials were read and used. Likely: credential harvesting from the standard AWS credentials location or `.env`, or prompt injection that caused an AWS API call. |
    • low Risky intent intent-offensive-security references/incident-response-playbook.md:201
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      | **username-password** | Login credentials were used. Likely: credential harvesting from password manager, `.env`, or agent memory. |
    • low Dangerous commands cmd-privilege references/security-compliance.md:237
      Privilege escalation / world-writable permissions (documentation of a security skill)
      sudo rm -f "$TRACEBIT_PATH" && echo "Removed CLI: $TRACEBIT_PATH" || echo "Could not remove CLI — run: sudo rm $TRACEBIT_PATH"
      security skill

    Files scanned: 13. Evidence is masked. Grey chips explain why severity was lowered.

    Against the Agent Skills spec

    ✓ No remarks against the Agent Skills spec

    Process rating: all ten parameters 64/100

    • 0Inputs and preconditions. Does not say what the process needs to start
    • 30Running it twice. 21 mutating operations with no state check
    • 40Result and completion. Does not say what the result is
    • 40Consistency. Frontmatter name (tracebit-canaries) differs from the folder (tracebit-canary-honeytokens)
    • 60Tools and files. Uses tools (bash, python) that frontmatter does not declare
    • 85Steps. 26 steps, 1 vague phrases
    • 100When it triggers. States when to use and when not to
    • 100Failures and branches. 1 branches, has a failure section
    • 100Execution cost. Instruction body is 2612 tokens
    • 100Progress reporting. Reports progress
    • medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
    • high The skill tells the model to perform an irreversible action with no human approval

    Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.

    Quality signals

    • +4Description does not say when NOT to use the skill (false activations)
    • +3Output format is not stated: the model decides each time
    • -32 of 4 scripts are never mentioned in SKILL.md
    • +1No license
    • +2Single-language instructions
    • +5Description quotes 2 example trigger phrases
    • +3Description length 546: enough signal without eating the budget
    • +4Structure: 16 headings
    • +3Step-by-step instructions: 26 items
    • +4Has examples (12 code blocks)
    • +4Reference files are cited in the instructions (6 of 6)

    Quality base 70; lint remarks subtract, signals add up to 100. Result: 90.

    External checks

    ClawHub: suspicious
    The skill has a legitimate canary-security purpose, but its broad activation and imperfect disclosures could let an agent start account setup, CLI authentication, credential-store changes, and recurring monitoring in situations where a user expected only security advice.
    LLM: suspicious (high) · 28 May 2026