benchmark

Benchmark mode marker — engagement objective is flag capture. Generic engagement rules apply unchanged.

Quality

55%

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/benchmark/SKILL.md

Quality

Content

71%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 content is highly actionable with executable bash and a concrete routing table, and is well-structured with clear sections. Its main weakness is workflow clarity on the batch flag sweep and short-circuit path, where a flag-verification checkpoint and retry loop are absent, capping that dimension at 3.

Suggestions

Add an explicit flag-verification checkpoint before the final re-echo (e.g. confirm the string matches the middleware-announced format) so the short-circuit has a validate→fix feedback loop.

Add a brief retry/fallback step for the sweep (e.g. if no path hits, fall back to the broad find) to turn the batch operation into a real loop.

Consider moving the 20-row tag→skill routing table into a separate reference file so the SKILL.md body stays a lean overview.

Dimension	Reasoning	Score
Conciseness	The body is efficient and assumes Claude's competence — flag-path sweep is a tight batched bash block and the routing table is a compact reference, with only minor explanatory padding (e.g. the 'What this skill is NOT' list and the rationale paragraphs for the routing table) that could be trimmed.	4 / 5
Actionability	Copy-paste-ready executable bash for the flag sweep and grep prefix extraction, plus an explicit numbered short-circuit procedure and a concrete tag→skill routing table — specific examples cover the common cases.	5 / 5
Workflow Clarity	The flag sweep is a batch operation with a guard ('do NOT issue one curl per path') and the short-circuit is sequenced, but there is no validation checkpoint confirming the flag is correct/verified before the final echo, and the destructive/batch sweep lacks a validate→fix→retry loop — the missing-feedback-loop cap of 3 applies.	3 / 5
Progressive Disclosure	Structure is good with clear section headers and mostly one-level-deep pointers to sibling SKILL.md files; no bundle files exist so references are external skill paths rather than a bundled reference tree, leaving minor organization gaps (e.g. the long routing table could live in a dedicated reference file).	4 / 5
	Total	16 / 20 Passed

Description

40%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 states the skill's niche clearly (benchmark flag-capture mode) but omits any explicit 'Use when...' trigger guidance and only weakly implies the invocation context. Specificity is moderate — one concrete behavior is named, but coverage is thin.

Suggestions

Add an explicit trigger clause, e.g. 'Use when running benchmark/CTF challenges where the objective is flag capture.'

Broaden trigger terms to include natural user phrases like 'CTF challenge' and 'capture the flag' alongside 'benchmark' and 'flag capture'.

Name a second concrete action (e.g. 'routes pre-declared vulnerability tags to exploit sub-skills') to lift specificity beyond a single behavior.

Dimension	Reasoning	Score
Specificity	Phrases 'engagement objective is flag capture' and 'Generic engagement rules apply unchanged' name the domain and one concrete behavior (flag capture), but no other concrete actions are enumerated — comparable to 'Processes PDF files and extracts content'.	3 / 5
Completeness	A 'what' is present (a mode marker for flag capture) but there is no 'Use when...' clause equivalent; per the judging guideline, a missing explicit trigger guidance caps completeness at 3, and here the 'when' is only weakly implied by 'mode marker'.	2 / 5
Trigger Term Quality	It surfaces 'benchmark', 'flag capture' as keywords but lacks the common synonyms a user would naturally say (e.g. 'CTF challenge', 'capture the flag'); several relevant variations are missing.	3 / 5
Distinctiveness Conflict Risk	'Benchmark mode marker' is a fairly distinct niche with a clear trigger that is unlikely to fire for unrelated skills, though the generic-sounding 'Generic engagement rules apply unchanged' leaves minor overlap risk with other engagement skills.	4 / 5
	Total	12 / 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/benchmark/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.