cicd

CI/CD pipeline attack category — poisoned pipeline execution, GitHub Actions expression injection, self-hosted runner abuse, secrets/OIDC exfil. Routing skill: fingerprint the CI provider + workflow surface, then load the matching leaf.

Quality

80%

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

Fix and improve this skill with Tessl

tessl review fix ./packages/decepticon/decepticon/skills/standard/exploit/cicd/SKILL.md

Quality

Content

90%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.

The body is a lean, highly actionable routing skill with copy-paste recon commands and a strong pre-active-test validation gate; its only weaknesses are a non-numbered recon workflow and substantial inlined triage content that nudges progressive disclosure and workflow clarity just below the top anchor.

Dimension	Reasoning	Score
Conciseness	The body is dense and information-rich — tables and copy-paste code blocks with no basic concept explanations; the two-sentence intro frames stakes without padding, assuming Claude's competence.	5 / 5
Actionability	Provides fully executable, copy-paste-ready commands — exact `gh api` calls, `grep -rE` patterns with precise regex, and `find`/`ls` recon — plus named tools (gato, actionlint, zizmor) covering the common cases.	5 / 5
Workflow Clarity	Sections flow logically (fingerprint → recon → triage → secret surface → decision gate) and the Decision gate is an explicit validation checklist before destructive active ops, satisfying the destructive-cap rule; held below 5 because recon steps are section-ordered rather than numbered and lack an error-recovery feedback loop.	4 / 5
Progressive Disclosure	Well-structured overview that routes to one-level-deep leaf skills via a clear 'When to Load' table and cross-references; held below 5 because substantial recon/triage content is inlined at the category level rather than split into referenced files.	4 / 5
	Total	18 / 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 specific and well-differentiated with strong natural trigger terms, but it omits an explicit 'Use when...' clause, leaving the activation condition only implied and capping completeness at 3.

Suggestions

Add an explicit 'Use when...' clause naming the activation triggers (e.g., 'Use when targeting CI/CD pipelines — GitHub Actions, GitLab CI, Jenkins, or other build systems — for pipeline injection, runner abuse, or secret/OIDC exfil').

Broaden provider trigger terms in the description itself (GitLab CI, Jenkins, CircleCI, Buildkite) rather than relying on the when_to_use metadata field.

Lead with a concrete action verb (e.g., 'Exploits CI/CD pipelines by...') so the capability reads as actions rather than a category list.

Dimension	Reasoning	Score
Specificity	Enumerates four concrete attack sub-categories ('poisoned pipeline execution, GitHub Actions expression injection, self-hosted runner abuse, secrets/OIDC exfil') plus a concrete routing action ('fingerprint the CI provider + workflow surface, then load the matching leaf'); held below 5 because the bulk is category nouns rather than the multiple action verbs in the 5 anchor.	4 / 5
Completeness	The 'what' is clear (CI/CD attack routing skill and its sub-categories) but there is no explicit 'Use when...' clause — the 'when' is only weakly implied, which the rubric caps at 3.	3 / 5
Trigger Term Quality	Natural terms a CI/CD security user would say are present ('CI/CD pipeline', 'GitHub Actions', 'self-hosted runner', 'OIDC'); provider synonyms like GitLab/Jenkins/CircleCI appear only in when_to_use metadata, not the description, so a few natural terms are missing.	4 / 5
Distinctiveness Conflict Risk	A clearly defined niche (CI/CD pipeline attacks) with distinct, specific triggers and a routing-to-leaf structure, giving minimal conflict risk with other skills.	5 / 5
	Total	16 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning

	Total	15 / 16 Passed

Repository: PurpleAILAB/Decepticon
Path: packages/decepticon/decepticon/skills/standard/exploit/cicd/SKILL.md
Commit: 0cf691e

Reviewed: 2 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.