CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

star-chamber

Multi-LLM craftsmanship council with live progress and debate mode for code review and design questions

1.29x
Quality

61%

Does it follow best practices?

Impact

87%

1.29x

Average score across 3 eval scenarios

SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./plugins/pragma/skills/star-chamber/SKILL.md

The canonical home for this skill is peteski22/star-chamber

SKILL.md
Quality
Evals
Security

Quality

Content

57%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, well-sectioned overview that correctly positions itself as a pointer document, with a concrete argument table and path setup. Its central weakness is that the single file it defers everything to — PROTOCOL.md, containing Steps 0-4, the CLI invocation, and debate mode — is not in the bundle, so the skill's executable guidance, validation checkpoints, and navigation chain all rest on a reference that cannot be verified.

Suggestions

Ship PROTOCOL.md with the skill bundle (or inline the essential Steps 0-4 and a sample CLI invocation) — the body defers the entire executable protocol to a file that is not present.

Add one concrete example of the CLI being invoked (flags, defaults, expected output) so the body alone demonstrates how the star-chamber binary is called.

Remove the duplicated invocation-mode explanation: the intro paragraph and the Auto-Invocation section both restate what 'model-invocable: false' in the frontmatter already says.

DimensionReasoningScore

Conciseness

The 34-line body is lean and sectioned, with a compact flag table and a minimal path-setup snippet. Not 5: there is redundant material that could be trimmed — the manual-vs-auto invocation split is stated twice (intro paragraph and the Auto-Invocation section, both restating what 'model-invocable: false' in the frontmatter already records), and the 'Manual-only flags' note re-explains the table's own 'Manual Only' column.

4 / 5

Actionability

The flag table (with defaults and repeatable semantics) and the executable path-variable snippet are concrete, but no example CLI invocation appears in the body, and the entire executable protocol lives in PROTOCOL.md — which is not present in the bundle. Not 4: the body alone offers no copy-paste-ready guidance for the skill's core action; the one concrete snippet only sets a shell variable.

3 / 5

Workflow Clarity

A sequence is present and named — path setup, then Steps 0-4 (prerequisite checks, mode detection, target identification, context gathering, CLI invocation, result presentation) — but each step is only a label; the substance and validation checkpoints are delegated to the absent PROTOCOL.md, leaving them implicit here. Not 4: anchor 4 expects the steps themselves to carry specific commands and checkpoints, not merely names.

3 / 5

Progressive Disclosure

Structurally the split is right: a concise overview with exactly one clearly signaled, one-level-deep reference ($STAR_CHAMBER_PATH/PROTOCOL.md, with the variable's origin explained in Path Setup). But that referenced file is absent from the bundle — no references/, scripts/, or assets/ directories exist — so the disclosure chain is unverifiable and effectively broken: every detail the skill promises lives in a file the reviewer cannot confirm. Not 4: good signaling cannot score well when the referenced target is missing.

3 / 5

Total

13

/

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 carves out a fairly distinct niche but reads as a noun-phrase stack: no action verbs, a metaphor ('craftsmanship council') where capabilities should be, and no 'Use when...' trigger guidance. It conveys the domain (multi-LLM, debate, code review, design questions) but not what the skill actually does — the fan-out and aggregation mechanics live only in the body.

Suggestions

Rewrite the description around concrete actions with an explicit trigger clause, e.g.: 'Fans out code review and design questions to multiple LLM providers (OpenAI, Gemini, etc.) and aggregates their feedback into consensus recommendations, with optional multi-round debate mode. Use when the user asks for a second opinion, multi-model review, or a debate on code or design decisions.'

Drop buzzwords and jargon from the description ('craftsmanship council', 'live progress') and instead name the providers and the deliverable (aggregated consensus feedback).

Add natural trigger keywords and synonyms such as 'second opinion', 'multi-model', or 'architecture feedback' so the skill surfaces for the phrasings users actually type.

DimensionReasoningScore

Specificity

The description is a noun-phrase stack — 'Multi-LLM craftsmanship council with live progress and debate mode' — with no action verbs and a metaphorical buzzword ('craftsmanship council'); it names the domain ('for code review and design questions') but the concrete mechanics (fan-out to providers, aggregation of feedback) appear only in the body, not the description. Not 3: anchor 3 requires 1-2 concrete actions, and this description states none — it relies on a metaphor plus feature labels ('live progress', 'debate mode').

2 / 5

Completeness

The 'what' is moderately clear — a multi-LLM council applied to code review and design questions — but the 'when' is entirely absent: there is no 'Use when...' clause or equivalent trigger guidance, which caps this dimension at 3 per the judging guidelines. Not 2: the 'what' is more than vague; the domain and mode of operation are identifiable.

3 / 5

Trigger Term Quality

'code review' and 'design questions' are natural terms a user would say, but common variations are missing ('second opinion', 'multi-model feedback', 'architecture review'), and 'craftsmanship council' and 'live progress' are internal jargon users would never say. Not 4: coverage stops at roughly two natural terms, short of 'good keyword coverage'.

3 / 5

Distinctiveness Conflict Risk

The multi-LLM council/debate niche is mostly distinct from ordinary single-model review skills; minor overlap risk remains with generic code-review skills via the phrase 'code review'. Not 5: without explicit triggers, 'code review' could still pull this skill for routine review requests it isn't meant to handle.

4 / 5

Total

12

/

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

allowed_tools_field

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

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

14

/

16

Passed

Repository

peteski22/agent-pragma

Path

plugins/pragma/skills/star-chamber/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.