CC cisa-cpg
Knowledge base from CISA Cross-Sector Cybersecurity Performance Goals 2.0. Use for prioritized IT/OT baseline cyber goals aligned to NIST CSF 2.0 (including GOVERN), critical-infrastructure hygiene, MSP/supply-chain expectations, and board-friendly outcome language. Covers CPG 2.0 goals only; does not replace full CSF programs, sector-specific goals (SSGs), or CSET tooling detail.
Knowledge base from CISA Cross-Sector Cybersecurity Performance Goals 2.0. Use for prioritized IT/OT baseline cyber goals aligned to NIST CSF 2.0 (including…
As a process C 58/100 · Has gaps — weak spots: result and completion, when it triggers, failures and branches
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 · 8
✓ No critical or high findings
Medium and low: 8
-
low Risky intent
intent-offensive-securitychapters/ch03-protect-goals.md:15Offensive-security / dual-use content (legitimate for authorised testing; review intended use) (detector / deny-list definition)- **3.E Monitor unsuccessful (automated) login attempts**: Detect password spraying and brute force early.
detector -
low Risky intent
intent-offensive-securitychapters/ch03-protect-goals.md:18Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- **3.I Implement logical/physical network segmentation**: Contain OT/IT blast radius and limit lateral movement.
-
low Risky intent
intent-offensive-securitychapters/ch03-protect-goals.md:23Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- **3.O Maintain system backups and restoration ability**: Ensure recoverable state against ransomware and destructive attacks.
-
low Risky intent
intent-offensive-securitychapters/ch03-protect-goals.md:40Offensive-security / dual-use content (legitimate for authorised testing; review intended use) (detector / deny-list definition)- **Backup targets reachable by the same ransomware identity**: Weakens 3.O.
detector -
low Risky intent
intent-offensive-securitychapters/ch05-it-ot-implementation-and-prioritization.md:16Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- **Prioritization method**: Use CISA/partner threat observations and your sector’s risk profile; do not treat the document order as a blind sequence if ransomware-ready backups (3.O) and MFA (3.F) ar
-
low Risky intent
intent-offensive-securitychapters/ch05-it-ot-implementation-and-prioritization.md:22Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- Start with identity, remote access, segmentation, backups, and logging — they unlock both IT ransomware resilience and OT remote-threat reduction.
-
low Risky intent
intent-offensive-securityglossary.md:14Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- **Network segmentation** — Logical/physical separation of networks to limit lateral movement and protect OT/IT zones. → ch03, ch05
-
low Risky intent
intent-offensive-securitypatterns.md:31Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- **Trade-offs:** Temporary access inconvenience vs ransomware initial access reduction.
Files scanned: 11. 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 58/100
- 0Result and completion. Does not say what the result is
- 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
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Tools and files. No external tools needed
- 100Steps. 36 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 1330 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)
- +3Output format is not stated: the model decides each time
- +4No input/output examples
- +2Single-language instructions
- +3Description length 383: enough signal without eating the budget
- +4Structure: 13 headings
- +3Step-by-step instructions: 36 items
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 81.