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 well-organized, action-oriented audit playbook with concrete grep/scanner/CLI commands and clear kill-signal checkpoints. Its main gaps are the incomplete Foundry POC template and reference files that are signaled but absent from the bundle.
Suggestions
Complete the Foundry POC template — replace '// ... buy tokens on Uniswap' and '// Assert token price crashed' placeholders with executable swap/assert calls so the examples are copy-paste ready.
Add an explicit feedback loop for findings: e.g. when a 'Kill if:' criterion fails, state the next action (flag severity, write finding to report, escalate to manual on-chain verification) to raise workflow clarity.
Either ship the referenced web3/*.md files in the bundle or note in SKILL.md that they are external/optional, so the one-level-deep references resolve to real content.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly lean — bulleted kill signals, copy-paste grep patterns, and terse 'Kill if:' criteria that assume Claude's competence; the Foundry POC template carries placeholder comments ('// ... buy tokens on Uniswap', '// Assert token price crashed') that could be trimmed or completed. Not a 5 because the POC skeleton pads the token budget with non-executable filler; not a 3 because everything else earns its place. | 4 / 5 |
Actionability | Provides concrete executable grep commands, scanner invocations, and Solana CLI checks that are copy-paste ready, but the Foundry POC template is a partial skeleton with '// ...' placeholders rather than fully executable code. Not a 5 due to those gaps; not a 3 because the majority of guidance is concrete and runnable. | 4 / 5 |
Workflow Clarity | Clear sequence — pre-dive kill signals → bug classes (each with a 'Kill if:' checkpoint) → automated scanner → POC → on-chain checks → references — with validation criteria and an explicit 'Scanner does NOT check' cross-check list. Not a 5 because there is no explicit error-recovery/feedback loop for findings that fail the kill criteria. | 4 / 5 |
Progressive Disclosure | Well-sectioned overview with three clearly signaled one-level-deep reference files at the end, each with a one-line description. Not a 5 because the referenced files (web3/10-meme-coin-bugs.md, web3/11-solana-token-audit.md, web3/12-dex-lp-attacks.md) are not present in the bundle, so navigation cannot be fully verified. | 4 / 5 |
Total | 16 / 20 Passed |