CtrlK
BlogDocsLog inGet started
Tessl Logo

fuzzing

Coverage-guided fuzzing methodology for compiled binaries and libraries: target scoping, fuzzer selection (AFL++/libFuzzer/Honggfuzz), harness skeleton, ASan+UBSan flags, corpus curation, crash triage and minimization (afl-tmin/minimize_corpus), exploitability rubric. Outputs corpora and crashes to /workspace/. Triggers on: 'fuzz', 'fuzzing', 'AFL', 'libFuzzer', 'harness', 'crash triage', 'coverage-guided', 'afl-tmin', 'sanitizer', 'corpus'.

74

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

85%Weight 40%Scale 1-5

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

A strong, executable playbook with concrete code for every phase of the fuzzing workflow and a validation step for crash evidence. Minor gains available in tightening inline comments and adding explicit verification checkpoints in the build/run phases.

Suggestions

Add an explicit 'verify the build is instrumented' checkpoint after section 3 (e.g., run a single input and confirm ASan is active) before proceeding to fuzzing.

Trim a few redundant inline code comments (e.g., '# replace with actual API', '# timeout per testcase (ms)') to improve token efficiency.

Consider moving the full ASan exploitability table and dedup script (sections 8–9) into a references/crash-triage.md file linked from the body, to keep SKILL.md as an overview.

DimensionReasoningScore

Conciseness

Mostly lean and code-driven with little explaining of concepts Claude already knows, though the intro line and a few prose notes ('5–50 seeds beats 5000 random bytes') are slightly more than minimal; a few inline comments in code could be trimmed.

4 / 5

Actionability

Provides copy-paste-ready bash and C code for building, harnessing, running AFL++/libFuzzer, minimizing corpora/crashes, and deduplicating crashes; concrete commands cover the common cases throughout.

5 / 5

Workflow Clarity

Steps are clearly sequenced (scoping → build → harness → corpus → run → triage → exploitability → persist/validate) with a validation checkpoint in section 10 (benign negative control + validate_workspace_finding), but several earlier steps (build, corpus, run) lack explicit verify-before-proceeding checkpoints.

4 / 5

Progressive Disclosure

Well-organized into numbered sections with a Tools Quick Reference table and no bundle files to over-split; content is appropriately self-contained, though the exploitability/triage detail could arguably live in a reference file and navigation between sections is implicit rather than cross-linked.

4 / 5

Total

17

/

20

Passed

Description

100%Weight 40%Scale 1-5

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

The description is highly specific, comprehensive, and explicit about both capabilities and trigger conditions, with strong natural keyword coverage. It is one of the longer descriptions but the detail is concrete rather than padded.

DimensionReasoningScore

Specificity

Lists multiple concrete actions across the full fuzzing lifecycle — target scoping, fuzzer selection, harness skeleton, ASan+UBSan flags, corpus curation, crash triage and minimization, and an exploitability rubric — with comprehensive coverage of the subdomain.

5 / 5

Completeness

Explicitly answers 'what' (coverage-guided fuzzing methodology with named steps) and 'when' via an explicit 'Triggers on:' clause listing concrete trigger phrases, matching the top anchor.

5 / 5

Trigger Term Quality

Includes natural user-facing terms and synonyms plus tool names ('fuzz', 'fuzzing', 'AFL', 'libFuzzer', 'harness', 'crash triage', 'coverage-guided', 'afl-tmin', 'sanitizer', 'corpus'), covering common phrasings a user would actually say.

5 / 5

Distinctiveness Conflict Risk

Occupies a clear niche (fuzzing compiled binaries/libraries) with distinct tool-specific triggers (AFL++/libFuzzer/Honggfuzz, afl-tmin) that minimize overlap with other skills.

5 / 5

Total

20

/

20

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

Total

15

/

16

Passed

Repository
PurpleAILAB/Decepticon
Reviewed

Table of Contents

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.