Content
75%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a highly actionable, well-sequenced live-collection runbook with strong per-OS command coverage and integrity checkpoints. Its main weakness is structure across files: none of the bundle's references, scripts, or the report template are referenced anywhere, so the detailed workflow material and automation the bundle provides cannot be found or used.
Suggestions
Add a 'References' or 'Bundle files' section near the top pointing to references/workflows.md (platform checklists and common pitfalls), references/standards.md (RFC 3227 / NIST mapping), references/api-reference.md, and assets/template.md (evidence report template).
Surface the automation path: mention scripts/process.py and scripts/agent.py (which already implement the ordered collection and hashing) as the preferred route, keeping the inline manual commands as a fallback for restricted hosts.
Add explicit validation feedback loops: instruct the responder to abort collection if the Step 1 toolkit hash diff fails, and to verify evidence hashes before and after any transfer (per the hashing protocol in references/workflows.md).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dominated by command blocks with no filler and every section earns its place, but the Key Concepts table briefly re-explains well-known principles (Locard's Exchange Principle, chain of custody) and some capture sets are redundant (netstat alongside ss and /proc/net/tcp) without stating the redundancy is intentional. | 4 / 5 |
Actionability | Every step provides concrete, copy-paste-ready commands with specific output paths, per-OS coverage (Windows/Linux/macOS), and an evidence manifest plus chain-of-custody record generated with real shell code — the common cases are fully covered. | 5 / 5 |
Workflow Clarity | The seven-step workflow follows a clear volatility-ordered sequence and includes real checkpoints (toolkit hash verification in Step 1, evidence hashing and custody documentation in Step 7). It falls short of a 5 because there are no feedback loops — e.g., what to do if the toolkit hash diff fails or a memory dump errors out — though checkpoints are present, keeping it above a 3. | 4 / 5 |
Progressive Disclosure | The body never mentions any bundle file: references/workflows.md (platform checklists and pitfalls), references/standards.md, references/api-reference.md, assets/template.md, and two ready-made collection scripts (scripts/process.py, scripts/agent.py) are all undiscoverable from SKILL.md, while ~170 lines of inlined platform commands duplicate workflows.md content — references are effectively buried. | 2 / 5 |
Total | 15 / 20 Passed |