CF optional-strict-instructions
Handle operations requiring user confirmation, permission verification, and strict adherence to explicit instructions. Activate when: (1) Operations need sudo/elevated permissions, (2) File deletion or modification is requested, (3) System changes are required, (4) User explicitly specifies a method (e.g., "use sudo"), (5) Sensitive operations need explicit consent, (6) Multiple valid approaches exist and user choice is required. Use this skill to ensure proper verification, user interaction, and strict compliance with user instructions.
As a process F 45/100 · Will not run — References files that are not bundled: scripts/template.sh
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 contains commands that delete files, rewrite disks or execute code fetched from the network. The agent may run them without asking if it believes the instructions require it.
Replace destructive commands with safe equivalents that ask for confirmation, scope them to one folder, and stop piping curl into a shell: pin a version and a checksum.
How to improve
- The text references files that are not there: add them or drop the references.
- 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 · 11
✓ No critical or high findings
Medium and low: 11
-
medium Dangerous commands
cmd-privilegereferences/scenarios.md:23Privilege escalation / world-writable permissionssudo rm -f "/home/user/important.pdf"
-
medium Dangerous commands
cmd-privilegereferences/scenarios.md:29Privilege escalation / world-writable permissions$ sudo rm -f "/home/user/important.pdf"
-
medium Dangerous commands
cmd-privilegereferences/scenarios.md:81Privilege escalation / world-writable permissionssudo cp /etc/nginx/nginx.conf /etc/nginx/nginx.conf.backup
-
medium Dangerous commands
cmd-privilegereferences/scenarios.md:91Privilege escalation / world-writable permissionssudo cp /etc/nginx/nginx.conf /etc/nginx/nginx.conf.backup.$(date +%s)
-
medium Dangerous commands
cmd-privilegeSKILL.md:96Privilege escalation / world-writable permissions1. sudo rm /var/log/app.log
-
medium Dangerous commands
cmd-privilegeSKILL.md:123Privilege escalation / world-writable permissions1. sudo rm -f (permanent, needs password)
-
low Dangerous commands
cmd-privilegescripts/strict-execution.sh:164Privilege escalation / world-writable permissions (string literal in code, not executed)"Use sudo rm -f (permanent deletion, requires password)"
code literal -
low Dangerous commands
cmd-privilegescripts/test-skill.sh:23Privilege escalation / world-writable permissions (test fixture / example file)1. sudo rm -f (permanent, needs password)
fixture -
low Dangerous commands
cmd-privilegescripts/test-skill.sh:39Privilege escalation / world-writable permissions (test fixture / example file)2. Execute: sudo rm test.txt
fixture -
low Dangerous commands
cmd-privilegescripts/test-skill.sh:46Privilege escalation / world-writable permissions (test fixture / example file)2. Execute: sudo rm test.txt
fixture -
low Dangerous commands
cmd-privilegeSKILL.md:82Privilege escalation / world-writable permissions (quoted — discussed, not commanded)2. Execute: `sudo rm file.txt`
quoted
Files scanned: 5. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
frontmatter-yamlSKILL.md: the frontmatter is not valid YAML (YAML parse error: Nested mappings are not allowed in compact mappings at line 2, column 14: description: Handle operations requiring user confirmation, permission verifica… ^ ); fields were read line by line. The usual cause is a colon inside an unquoted value - warning
missing-refreference to a missing file: scripts/template.sh
Process rating: all ten parameters 45/100
- 0Tools and files. 1 referenced file(s) missing: scripts/template.sh
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 2 mutating operations with no state check
- 100Steps. 30 steps
- 100Failures and branches. 2 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 1336 tokens
- 100Progress reporting. Reports progress
- 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
- +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
- -32 of 2 scripts are never mentioned in SKILL.md
- +1No license
- +2Single-language instructions
- +3Description length 543: enough signal without eating the budget
- +4Structure: 31 headings
- +3Step-by-step instructions: 30 items
- +4Has examples (12 code blocks)
- +4Reference files are cited in the instructions (1 of 1)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 69.