CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

benchmark

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

Quality

68%

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/benchmark/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

75%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 well-structured mode-marker skill with executable sweep commands, an explicit short-circuit protocol, and a dense but justified routing table. The main gaps are the assumed query-param RCE sink shape in the sweep code and an unspecified flag-verification procedure.

DimensionReasoningScore

Conciseness

The body is lean — a dense routing table, copy-paste bash sweeps, and no explanation of concepts Claude already knows — with only minor asides that could be trimmed (e.g. 'Recon does NOT consume this table — recon's role is observation' and the benchmark-vs-real-engagement rationale paragraph).

4 / 5

Actionability

The flag-path sweep is fully written bash with explicit placeholder-substitution guidance ('Replace `<RCE_SINK>` with the confirmed injection endpoint') and the short-circuit steps are concrete and numbered, but the sweep hardcodes a `?cmd=` query-param sink shape that may not match every confirmed RCE primitive.

4 / 5

Workflow Clarity

Sequencing is explicit (RCE confirmed → flag sweep → credential sweep → verbatim final echo) and the batch sweep includes verification steps (non-empty result check, grep for the announced flag format), but the short-circuit requires a 'verified flag' without specifying how to verify one.

4 / 5

Progressive Disclosure

Sections are well-organized, the 'What this skill is NOT' section aids navigation, and all external skill references are one level deep and clearly signaled; the 21-row routing table is bulky inline, though the body justifies it as the canonical single home for the mapping.

4 / 5

Total

16

/

20

Passed

Description

48%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 concise and distinctive but undersells the skill: it names the mode and objective yet omits any trigger guidance and any of the skill's concrete behaviors. Much of the useful trigger vocabulary (ctf, challenge) is stranded in metadata instead of the description.

Suggestions

Add an explicit 'Use when...' clause to the description, e.g. 'Use when running in benchmark/CTF mode or when the engagement objective is flag capture', so completeness is not capped at 3.

Fold the natural trigger terms from metadata.when_to_use (benchmark, ctf, challenge, flag capture) into the description itself so users searching those words surface the skill.

Name 1-2 concrete behaviors in the description — e.g. flag-path sweep after RCE and tag-to-skill routing — so the specificity rises above a bare domain label.

DimensionReasoningScore

Specificity

The description names the domain ('Benchmark mode marker') and one concrete fact ('engagement objective is flag capture') but lists no concrete actions or capabilities the skill provides, matching 'Names the domain but actions are minimal or generic'.

2 / 5

Completeness

The 'what' is clear (marks benchmark mode, sets the objective to flag capture, defers to generic rules) but there is no 'Use when...' clause or equivalent explicit trigger guidance, capping completeness at 3 per the judging guidelines.

3 / 5

Trigger Term Quality

'benchmark' and 'flag capture' are natural terms a user might say, but common variations and synonyms (CTF, challenge, flag format) appear only in metadata.when_to_use and are missing from the description itself.

3 / 5

Distinctiveness Conflict Risk

'Benchmark mode' plus 'flag capture' forms a clear niche that is mostly distinct from generic engagement skills and explicitly defers to them ('Generic engagement rules apply unchanged'), with only minor overlap risk against general CTF-oriented skills.

4 / 5

Total

12

/

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

metadata_version

'metadata.version' is missing

Warning

Total

15

/

16

Passed

Repository

PurpleAILAB/Decepticon

Path

packages/decepticon/decepticon/skills/benchmark/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.