CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

malware-triage

Triage a suspicious file statically and safely, and turn it into something defenders can use: a verdict with confidence and evidence, defanged indicators as CSV or STIX 2.1, a YARA signature tested against goodware, and ATT&CK-mapped capabilities to hand to detection and hunting. Covers safe handling, sample identity (SHA-256, ssdeep, TLSH, imphash), reputation lookup without uploading, file type, entropy and packers, strings and FLOSS, PE / ELF / Mach-O metadata, script and Office macro deobfuscation, archives, LNK and ISO, and capa. Use when the user mentions 'malware triage,' 'malware analysis,' 'analyze this sample,' 'suspicious file,' 'is this file malicious,' 'suspicious attachment,' 'malicious macro,' 'maldoc,' 'olevba,' 'capa,' 'FLOSS,' 'packed binary,' 'imphash,' 'YARA,' 'VirusTotal,' 'defang,' 'STIX,' or has a binary, script or document they want looked at without running it.

Quality

94%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

92%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 mature, highly actionable skill: executable commands for every format family, a validated YARA workflow, a calibrated verdict rubric, and clean one-level-deep progressive disclosure into three real reference files. The only cost is length — the intro and a few meta-asides could be trimmed without losing any instructional value.

Suggestions

Trim the opening paragraph and the cross-references paragraph, which largely restate the description's scope and trigger content — the highest-token, lowest-instruction text in the body.

Compress the 'Why not findings.json?' aside to a single sentence (or fold it into the report-format section), since it justifies a format choice rather than instructing the analyst.

Consider moving the ten-item external bibliography to a reference file — it is rarely consulted mid-execution and would shorten the always-loaded body.

DimensionReasoningScore

Conciseness

The body is dense and assumes Claude's competence — it never explains what a hash, YARA, or a PE file is, and every judgment call ('Packed is not malicious', 'Signed is not benign', 'No hit is not clean') is domain-specific knowledge worth its tokens. Minor trims remain: the opening paragraph restates the description nearly verbatim, the cross-references paragraph is long, and the 'Why not findings.json?' aside justifies rather than instructs, so it matches anchor 4 ('efficient; minor instances that could be trimmed') rather than anchor 5's 'every token earns its place'.

4 / 5

Actionability

Copy-paste-ready commands with flags throughout: 'shasum -a 256', 'strings -a -n 6 -e l', 'floss -n 6 -- sample.bin', 'capa -j sample.bin', 'olevba --decode --reveal', the pefile imphash one-liner, ready-made rg/grep triage patterns, a populated CSV example, and a full report template. Common cases (PE, ELF, Mach-O, scripts, Office, archives/LNK/ISO) are each covered with the concrete tool invocation, matching 'fully executable; specific examples cover the common cases'.

5 / 5

Workflow Clarity

A clearly sequenced 10-step flow with explicit validation checkpoints: hash 'before anything else', YARA rules must 'match, and show which strings hit' with 'must be 0' on a goodware corpus ('A rule nobody tested against goodware is a draft'), and the verdict step has an evidence bar plus a devil's-advocate check ('argue the other side once') with four worked cases. Recovery paths are stated (re-run Steps 4–6 on the unpacked copy; unknown verdict → 'say what would resolve it'), matching anchor 5's explicit validation and feedback loops.

5 / 5

Progressive Disclosure

The body is an actionable overview that links three existing one-level-deep reference files exactly at point of need: static-triage.md from Steps 3 and 5, ioc-export.md from Step 7, yara-rules.md from Step 8 ('Full guidance is in [references/yara-rules.md]'). All three files exist in references/ with no nesting, and detail (per-format commands, STIX field lists, YARA style guidance) is appropriately deferred to them, matching anchor 5.

5 / 5

Total

19

/

20

Passed

Description

96%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.

An excellent description: dense with concrete deliverables, a genuine trigger list of phrases real users say, and an explicit Use-when clause. The only blemish is that the broadest triggers ('YARA', 'malware analysis') could pull in requests better served by dedicated YARA-authoring or full-analysis skills.

DimensionReasoningScore

Specificity

Lists multiple concrete deliverables — 'a verdict with confidence and evidence, defanged indicators as CSV or STIX 2.1, a YARA signature tested against goodware, and ATT&CK-mapped capabilities' — plus a detailed coverage list (hashing, entropy and packers, PE/ELF/Mach-O metadata, macro deobfuscation, capa). This matches the anchor 'multiple specific concrete actions; comprehensive coverage'; nothing is vague or padded.

5 / 5

Completeness

Explicitly answers both: the 'what' (verdict with confidence and evidence, defanged indicators, tested YARA rule, ATT&CK-mapped capabilities) and an explicit 'Use when the user mentions ...' clause with concrete trigger phrases. This is a direct match for the anchor-5 example structure; not anchor 4, where the 'when' is present but less specific.

5 / 5

Trigger Term Quality

Eighteen natural trigger phrases users would actually say, including synonyms ('maldoc', 'malicious macro', 'suspicious attachment') and a full question form ('is this file malicious'), plus tool names (olevba, capa, FLOSS, YARA, VirusTotal). Coverage is comprehensive across synonyms and phrasings; only file extensions (.docm, .exe, .lnk) are absent, which keeps it at anchor 5 rather than above it, matching 'comprehensive coverage of natural terms' more closely than anchor 4's 'a few natural terms missing'.

5 / 5

Distinctiveness Conflict Risk

The framing 'statically and safely' / 'looked at without running it' carves a clear niche distinct from detonation and reverse-engineering skills, and most triggers (maldoc, imphash, defang, STIX) are distinctive. However, the broad triggers 'YARA' and 'malware analysis' create minor overlap risk with a dedicated YARA-authoring or full-analysis skill, which fits anchor 4 ('mostly distinct; minor overlap risk with closely related skills') better than anchor 5's 'minimal conflict risk'.

4 / 5

Total

19

/

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.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

allowed_tools_field

'allowed-tools' contains unusual tool name(s)

Warning

Total

15

/

16

Passed

Repository

briiirussell/cybersecurity-skills

Path

skills/malware-triage/SKILL.md

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.