CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ti-ioc-extraction

Automated IOC extraction from threat reports, logs, and unstructured text — parse hashes, IPs, domains, URLs, email addresses, and CVEs. Covers regex-based extraction, defanging/refanging, bulk hash lookup, IOC deduplication, YARA rule generation from IOCs, and STIX/TAXII formatting for sharing.

Quality

72%

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

Fix and improve this skill with Tessl

tessl review fix ./packages/decepticon/decepticon/skills/standard/analyst/ti-ioc-extraction/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

65%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 dense, highly executable reference with real code for every stage of the pipeline (extract → refang → dedup → look up → share), plus genuinely useful error-handling and routing tables. Weaknesses are the missing validation checkpoints before batch/external-push operations, the monolithic single-file layout, and a few fragile code spots.

Suggestions

Add validation checkpoints around the batch steps: verify deduped.json counts (and spot-check a few indicators) before building the STIX bundle or pushing to MISP, and remove 2>/dev/null from the VirusTotal/AbuseIPDB loops so lookup failures surface instead of being silenced.

Move the large Python pipelines (dedup, STIX bundle builder, MISP push) into scripts/ files and reference them from SKILL.md, keeping the quick reference, decision gate, and tables inline.

Fix the MISP push snippet: use event.add_attribute(type=misp_type, value=item, to_ids=True) rather than add_attribute(**attr) on a MISPAttribute, which will raise at runtime.

DimensionReasoningScore

Conciseness

The body is code-dominant and lean in prose — sections open directly with executable commands and assume Claude's competence, with almost no explanation of known concepts. Minor trims are available: the Quick Reference repeats Section 1's md5/sha256/IPv4 regexes verbatim, and the Detection Signatures table restates Section 1's patterns plus well-known facts (MD5 = 32 hex chars, CVE format). Efficient with minor redundancy matches anchor 4; it is not anchor 5 because of that duplication, and not anchor 3 because there is no verbose conceptual padding.

4 / 5

Actionability

Nearly everything is copy-paste ready: grep/sed one-liners, complete Python heredocs for dedup/STIX/MISP, curl loops for VirusTotal/MalwareBazaar/AbuseIPDB, and CyberChef recipes. Minor gaps keep it below anchor 5: the MISP snippet calls event.add_attribute(**attr) on a MISPAttribute object (not a mapping, so it will raise at runtime), and the bulk lookup loops pipe through 2>/dev/null, silently swallowing failures.

4 / 5

Workflow Clarity

The Decision Gate provides useful routing (PDF → pdftotext first; paste dump → refang first; >100 hashes → MalwareBazaar batch; ioc-finder unavailable → Section 1 regex fallback) and the Error Handling table aids recovery, but the batch and external-push operations have no validation checkpoints — nothing verifies extracted/deduped IOC counts before building the STIX bundle or pushing to MISP, and lookup errors are suppressed. Per the rubric guideline, batch operations without validation cap workflow_clarity at 3.

3 / 5

Progressive Disclosure

Sections are clearly organized with headers, tables, and a decision gate, but the skill is a single ~425-line SKILL.md with no bundle files: the full dedup, STIX, and MISP Python pipelines and the entire regex library are inlined where the format supports scripts/ and references/ files. This is 'content that should be separate is inline' (anchor 3), and the under-50-line simple-skill exception clearly does not apply.

3 / 5

Total

14

/

20

Passed

Description

70%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 concrete, tool-aware, and rich in natural trigger vocabulary, defining a distinct IOC-extraction niche. Its main gaps are the missing 'Use when...' trigger clause (capping completeness at 3) and an over-claimed YARA capability that the skill body does not back.

Suggestions

Append an explicit trigger clause to the description, e.g. 'Use when the user asks to extract IOCs or indicators of compromise from a threat report, defang/refang indicators, run bulk hash lookups, or format IOCs as STIX for sharing.'

Remove 'YARA rule generation from IOCs' from the description, or add a matching YARA section to the body — the claim is currently unsupported and counts as an over-claim.

Fold one or two high-value synonyms ('indicator of compromise', 'threat intelligence', 'MISP') into the description text itself rather than relying solely on metadata.when_to_use.

DimensionReasoningScore

Specificity

The description lists many concrete actions — 'parse hashes, IPs, domains, URLs, email addresses, and CVEs', 'regex-based extraction, defanging/refanging, bulk hash lookup, IOC deduplication, ... STIX/TAXII formatting' — approaching the comprehensive anchor 5, but 'YARA rule generation from IOCs' is an over-claim: the skill body contains no YARA guidance at all, and the guidelines direct penalizing over-claims. It is well above anchor 3 (only 1-2 concrete actions).

4 / 5

Completeness

The 'what' is clear and comprehensive, but the description field contains no 'Use when...' clause or equivalent trigger guidance of its own, which caps completeness at 3 per the judging guideline. It is not a 4 because 'when' is absent rather than merely implicit, and not a 2 because the 'what' is fully explicit rather than vague.

3 / 5

Trigger Term Quality

Good coverage of natural phrases users would say — 'IOC extraction', 'threat reports', 'parse hashes', 'defanging/refanging', 'bulk hash lookup', 'STIX/TAXII'. A few natural terms are missing from the description field itself ('indicator of compromise' spelled out, 'MISP', 'CyberChef', 'threat intelligence'), which the frontmatter's metadata.when_to_use field supplies instead — matching anchor 4 rather than anchor 5's comprehensive synonym coverage.

4 / 5

Distinctiveness Conflict Risk

It carves a clear niche — automated IOC extraction from threat text with distinctive vocabulary ('defanging/refanging', 'bulk hash lookup', 'IOC deduplication', 'STIX/TAXII', 'YARA') — that is unlikely to collide with adjacent skills. This matches anchor 5 (clear niche, distinct triggers, minimal conflict risk).

5 / 5

Total

16

/

20

Passed

Validation

87%

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

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

relative_links

Relative link issues: 1 missing

Warning

Total

14

/

16

Passed

Repository

PurpleAILAB/Decepticon

Path

packages/decepticon/decepticon/skills/standard/analyst/ti-ioc-extraction/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.