BC quorum
Multi-agent validation framework — 6 independent AI critics evaluate artifacts against rubrics with evidence-grounded findings.
As a process C 52/100 · Has gaps — weak spots: when it triggers, inputs and preconditions, failures and branches
AnalyzerGitHubAI and agentsData and analyticstype and topics are labelled automatically from the skill text
How to improve
- Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
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 · 14
✓ No critical or high findings
Medium and low: 14
-
low Risky intent
intent-offensive-securitydocs/critics/SEC03_EVIDENCE_INTEGRITY.md:31Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| **AI red team** | Test system integrity | High | Systematically probes for minimal artifacts that produce PASS verdicts |
-
low Risky intent
intent-offensive-securitydocs/critics/SECURITY_CRITIC_FRAMEWORK.md:205Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| Privilege escalation paths | Is there any code path where a lower-privilege user can trigger actions reserved for higher-privilege users? Horizontal (same role, different user) and vertical (lower r
-
low Risky intent
intent-offensive-securitydocs/critics/SECURITY_CRITIC_FRAMEWORK.md:644Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| PS-18 | COM object abuse for privilege escalation | *(no PSSA rule)* | CWE-269, ASVS V8 | T3 (LLM) |
-
low Risky intent
intent-offensive-securitydocs/critics/SECURITY_CRITIC_FRAMEWORK.md:667Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| Privilege escalation | ❌ None | ✅ Good | **LLM-ONLY** |
-
low Risky intent
intent-offensive-securitydocs/critics/SECURITY_CRITIC_FRAMEWORK.md:707Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| Evaluate authorization logic — IDOR, privilege escalation, missing auth checks | SEC-04 | CWE-862, CWE-863, ASVS V8 |
-
low Risky intent
intent-offensive-securitydocs/research/security-code-review-frameworks.md:50Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| **V8** | Authorization | Access control, IDOR, privilege escalation, RBAC/ABAC | ✅ Yes | ✅ High |
-
low Dangerous commands
cmd-eval-dynamicdocs/research/security-code-review-frameworks.md:135Dynamic code execution from decoded/untrusted input (documentation table row; documentation of a security skill)| CWE-78 | `os.system()`, `subprocess` without shell=False | `os.system(f"ping {host}")` |tablesecurity skill -
low Risky intent
intent-offensive-securitydocs/research/security-code-review-frameworks.md:305Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| **SA-11(5)** | Penetration Testing | Pen test with defined breadth/depth/constraints | Low — runtime testing, not code review |
-
low Risky intent
intent-offensive-securitydocs/research/security-code-review-frameworks.md:373Offensive-security / dual-use content (legitimate for authorised testing; review intended use)| Privilege escalation paths | ❌ None | ✅ Good | Requires understanding permission model |
-
low Secrets in code
secret-anthropic-keyreference-implementation/examples/sample-agent-config.yaml:71Anthropic API key (placeholder value)anthropic_api_key: sk-a…key
placeholder
A further 4 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 53. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
description-no-whendescription does not say WHEN to use the skill (no "use when")
Process rating: all ten parameters 52/100
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Failures and branches. Linear process with no failure handling
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 1 mutating operations with no state check
- 60Tools and files. Uses tools (python) that frontmatter does not declare
- 60Result and completion. Output format stated, no completion criterion
- 100Steps. 14 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 968 tokens
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
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- -44 reference files, but SKILL.md never points to them: the model will not open them
- +1No license
- +2Single-language instructions
- +3Description length 127: enough signal without eating the budget
- +4Structure: 8 headings
- +3Step-by-step instructions: 14 items
- +3Output format is stated explicitly
- +4Has examples (5 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 71.
External checks
ClawHub: clean
Quorum is a disclosed validation tool, but users should treat it as a file-reading LLM workflow and not run it on private projects without understanding where analyzed content is sent.
LLM: benign (medium) · VirusTotal: benign · 28 May 2026