AB permission-guard
Security watchdog for OpenClaw agents that monitors installed skill behavior, detects unauthorized file access, suspicious outbound network calls, dangerous command patterns, and generates permission audit reports. Use this skill whenever the user asks about agent activity ("what did my agent do", "check what my skill accessed", "monitor agent permissions", "permission report", "activity log", "did my agent do anything weird", "skill behavior audit", "what files did my agent touch"), after installing a new skill to establish a behavior baseline, or when suspicious or unexpected behavior is suspected. Trigger proactively after any skill installation — a first-run baseline check is always worthwhile.
As a process B 65/100 · Nearly there — weak spots: inputs and preconditions, consistency, running it twice
What is at stake
Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
The skill asks for more than the task needs: broad tool access, credential environment variables, binaries. Every extra permission widens the damage from a mistake or a compromise.
Narrow allowed-tools and the variable list to the minimum; replace binaries with readable sources or scripts.
How to improve
- 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 · 1
✓ No critical or high findings
Medium and low: 1
-
medium Broad scope
meta-broad-allowed-toolsSKILL.md:1Broad tool permissions pre-approved: Bashallowed-tools: Bash Read Write
Files scanned: 2. 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 65/100
- 0Inputs and preconditions. Does not say what the process needs to start
- 30Running it twice. 1 mutating operations with no state check
- 40Consistency. Frontmatter name (permission-guard) differs from the folder (permission-guard-v1)
- 55Failures and branches. 1 branches
- 60Result and completion. Output format stated, no completion criterion
- 70When it triggers. States when to use, but not when not to
- 85Steps. 11 steps, 1 vague phrases
- 100Tools and files. Tools declared in frontmatter
- 100Execution cost. Instruction body is 1224 tokens
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
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)
- +1No license
- +2Single-language instructions
- +5Description quotes 8 example trigger phrases
- +3Description length 707: enough signal without eating the budget
- +4Structure: 11 headings
- +3Step-by-step instructions: 11 items
- +3Output format is stated explicitly
- +4Has examples (6 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 92.