SKILLEMALL.ai
GitHub ↗ RU Sign in

BF black-fortress

Pre-installation agentic sandboxing protocol. 5-layer defense: semantic neutralization, hard quarantine, kernel ground-truth, trusted output rendering, and sterile autopsy.

ClawHub Hermes author: Jahfali-dev v1.1.10 MIT-0 13 files body ≈ 5 766 tokens Open the sourceclawhub.ai↗ analyzed 3 d ago

As a process F 47/100 · Will not run — References files that are not bundled: LICENSE

GeneratorDockerInfrastructureAI and agentstype and topics are labelled automatically from the skill text
Runs in: OpenClaw Hermes Agent
JSON
Technical rating
B
80/100
safety, quality, tests
Safety 60%
97
Quality 40%
55
Run on models
none yet
Process rating
F
47/100
Will not run
References files that are not bundled: LICENSE
Tools and files w 18
0
Result and completion w 14
0
When it triggers w 12
20
the three weakest of ten parameters · all ten

How to improve

  1. For Hermes the description must be one sentence under 60 characters; move the conditions to a "When to Use" section.
  2. The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
  3. The text references files that are not there: add them or drop the references.
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 · 3

✓ No critical or high findings

Medium and low: 3
  • low Dangerous commands cmd-privilege scripts/behavioral_audit.py:51
    Privilege escalation / world-writable permissions (string literal in code, not executed; documentation of a security skill)
    "setuid", "setgid", "setreuid", "setregid",
    code literalsecurity skill
  • low Risky intent intent-offensive-security SKILL.md:144
    Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
    | `/bin/sh` / `/bin/bash` | ❌ | No reverse shell, no `os.system()` |
  • low Risky intent intent-offensive-security SKILL.md:149
    Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
    | UID 0 (root) | ❌ | No privilege escalation |

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

Against the Agent Skills spec

  • warning description-long-hermes description is 172 chars; the Hermes authoring standard requires ≤ 60 (one sentence, ending with a period)
  • warning body-long SKILL.md body ≈ 5766 tokens (recommended < 5000); move details to references/
  • warning missing-ref reference to a missing file: LICENSE
  • note frontmatter-key unknown frontmatter key "required_commands"
  • note frontmatter-key unknown frontmatter key "required_environment_variables"
  • note frontmatter-key unknown frontmatter key "required_privileges"
  • note frontmatter-key unknown frontmatter key "docker_image"
  • note frontmatter-key unknown frontmatter key "docker_image_source"

Process rating: all ten parameters 47/100

Will not run. References files that are not bundled: LICENSE
  • 0Tools and files. 1 referenced file(s) missing: LICENSE
  • 0Result and completion. Does not say what the result is
  • 20When it triggers. No condition that starts the skill
  • 30Running it twice. 15 mutating operations with no state check
  • 60Consistency. The Hermes dialect needs category and tags
  • 70Inputs and preconditions. Inputs and preconditions are listed
  • 70Execution cost. Instruction body is 5766 tokens
  • 100Steps. 74 steps
  • 100Failures and branches. 5 branches, has a failure section
  • 100Progress reporting. Reports progress
  • medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
  • low 16 top-level sections: this looks like several domains in one skill

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)
  • +3Output format is not stated: the model decides each time
  • -215 emoji in the instructions: noise for the model
  • -32 of 9 scripts are never mentioned in SKILL.md
  • +2Single-language instructions
  • +3Description length 172: enough signal without eating the budget
  • +4Structure: 41 headings
  • +3Step-by-step instructions: 74 items
  • +4Has examples (18 code blocks)
  • +1License stated

Quality base 70; lint remarks subtract, signals add up to 100. Result: 55.

External checks

ClawHub: suspicious
This is a defensive local sandboxing skill, but it overstates important protections and asks users to trust high-impact Docker/root workflows that are not fully implemented or scoped.
LLM: suspicious (high) · VirusTotal: suspicious · 28 May 2026