reverse-engineering

Provides reverse engineering techniques. Use when the main job is to understand how a compiled, obfuscated, packed, or virtualized target works before exploiting or solving it, including binaries, APKs, WASM, firmware, custom VMs, bytecode, malware-like loaders, and anti-debug or anti-analysis logic. Do not use it when the vulnerability is already understood and the remaining task is exploitation; use pwn instead. Do not use it for pure web workflows, log or disk forensics, or standalone crypto problems unless reversing the implementation is the real blocker.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Quality

Content

77%Weight 40%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A highly actionable, well-sequenced RE reference with concrete commands and a validation checklist, but it is padded with inlined tool enumerations and references a large library of bundle files that are not actually present in the bundle, undermining progressive disclosure.

Suggestions

Trim the 'Additional Resources' bullets to one-line pointers and move the exhaustive parenthetical tool lists into the referenced files so the overview stays lean.

Reconcile reference paths with the actual bundle: either add the missing files (tools.md, tools-dynamic.md, anti-analysis.md, patterns*.md, languages*.md, field-notes.md, etc.) to ./references/ or remove/correct the broken links and external ../ paths (../field-journal/precedent-reverse.md, ../tool-index.md, ida-reverse/, apk-reverse/).

Reduce the bilingual operational scaffolding (ACTION REQUIRED, 路由上下文, 任务完成自检) to the minimal gating needed, or relocate it to a referenced workflow file, to reclaim context tokens.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient with copy-paste-ready command blocks, but the 'Additional Resources' section inlines very long parenthetical tool enumerations (e.g. the tools-dynamic.md bullet listing Frida/angr/llbd/x64dbg/Qiling/Triton inline) that duplicate content belonging in the bundle files, and the Chinese operational scaffolding (ACTION REQUIRED, 路由上下文, 任务完成自检) adds tokens.	2 / 3
Actionability	It provides fully executable, copy-paste-ready commands throughout: install snippets ("pip install frida-tools angr qiling...", "apt install gdb radare2..."), quick wins ("strings binary \| grep -E 'flag\{\|CTF\{'", "ltrace ./binary"), and specific GDB techniques ("b *main+0xca", "x/s $rsi").	3 / 3
Workflow Clarity	A clear sequenced 'Problem-Solving Workflow' (strings -> ltrace -> Frida -> angr -> Qiling -> map control flow -> automate) is paired with explicit validation: step 8 'Validate assumptions by comparing decompiler outputs' and a terminal self-check checklist marked 'MUST 通过' before claiming completion, plus staged NOW/NEXT/ACT gating.	3 / 3
Progressive Disclosure	The SKILL.md is structured as an overview pointing to separate files (good intent, one-level-deep), but scored against the actual bundle only 3 reference files exist (re-agent-workflow.md, ollvm-deobfuscation.md, ai-assisted-re.md) while the body references ~14 missing files (tools.md, tools-dynamic.md, patterns.md, languages.md, field-notes.md) plus external ../ paths, so navigation is largely broken and excessive detail is inlined rather than split out.	2 / 3
	Total	10 / 12 Passed

Description

90%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, third-person description with an explicit 'Use when' clause, concrete target enumeration, and clear negative-routing against pwn/web/forensics/crypto. The only weakness is that the stated capability is a single action family rather than multiple distinct actions.

Dimension	Reasoning	Score
Specificity	The description names the domain and concrete target types ("compiled, obfuscated, packed, or virtualized target"; "binaries, APKs, WASM, firmware, custom VMs, bytecode, malware-like loaders"), but the action itself is essentially a single capability family ("Provides reverse engineering techniques" / "understand how a ... target works") rather than multiple distinct concrete actions like the score-3 anchor's extract/fill/merge.	2 / 3
Completeness	It explicitly answers both what ("Provides reverse engineering techniques ... understand how a ... target works") and when ("Use when the main job is to understand how a compiled, obfuscated, packed, or virtualized target works"), with explicit "Do not use it when ..." triggers as well.	3 / 3
Trigger Term Quality	It surfaces natural terms a user would say ("reverse engineering", "binaries", "APKs", "WASM", "firmware", "packed", "obfuscated", "bytecode") with good coverage, rather than pure technical jargon.	3 / 3
Distinctiveness Conflict Risk	It carves a clear niche with explicit negative routing ("use pwn instead"; "Do not use it for pure web workflows, log or disk forensics, or standalone crypto problems unless reversing the implementation is the real blocker"), making conflict with adjacent skills unlikely.	3 / 3
	Total	11 / 12 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
relative_links	Relative link issues: 15 missing	Warning

	Total	14 / 16 Passed

Repository: zhaoxuya520/reverse-skill
Path: skills/reverse-engineering/SKILL.md
Commit: 6aa1362

Reviewed: 7 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.