CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ce-pov

Judge a supplied subject against the project's evidence and constraints. Use when assessing an external-adoption question, a holistic take on a document, or a supplied approach set. Use for an oracle panel to consult other models and reconcile their opinions. Use ce-explain for understanding and ce-doc-review for findings review.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Critical

Do not install without reviewing

SKILL.md
Quality
Evals
Security

Quality

Content

92%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.

An exemplary orchestration skill body: a tight phase-gated workflow with verification checkpoints and explicit failure paths, lean token-efficient prose, and well-signaled one-level-deep references verified to exist. The only dimension below 5 is actionability, where a few body-level directives stay abstract because the executable mechanics are intentionally delegated to reference files.

DimensionReasoningScore

Conciseness

The body is lean and directive throughout — no concept explanations Claude already knows, no padding, and every sentence issues an instruction or boundary ('Stay read-only while forming and reconciling the POV', 'Nothing the conversation asserts substitutes for grounding'). Matches the anchor 'every token earns its place'; not 4 because there is no over-explanation to trim, only earned boundary-drawing.

5 / 5

Actionability

Gives concrete directives at each phase ('Read `references/intake.md` now', 'return **Blocked — missing context**', 'Read `docs_root` from `<repo-root>/.compound-engineering/config.yaml`', 'Send scouts directly to candidate-specific current evidence') with an explicit blocked path. Not 5 because a few directives remain abstract at the body level ('Deliver a supported position in the form the intended consumer can use') with the real mechanics delegated to references; not 3 because nothing is pseudocode or missing key details — each step states exactly what to do next.

4 / 5

Workflow Clarity

Explicit phase sequence (Phase 0 Frame → 1 Ground → 2 Verify Grounding → 3 POV → 4 Deliver) with read-gates between phases, a dedicated verification checkpoint ('Apply that check over the grounded evidence. If the evidence falls short, no subject shape may return a confident result'), and defined failure recovery (Blocked — missing context, panel-branch-before-result ordering). Matches the anchor for clear sequence with explicit validation steps and error-recovery handling; this is a read-only judgment skill, so the destructive/batch cap does not apply.

5 / 5

Progressive Disclosure

The body is a concise orchestration overview that signals six reference files at the exact point of need with bold 'Read X now' directives; all referenced files (intake.md, boundaries.md, grounding.md, method.md, cross-model-panel.md, followup.md) verified to exist, and heavy content (37KB cross-model-panel.md, 109KB peer-job-runner.py) is split out of SKILL.md. Cross-links among references point only to siblings the body also links directly, so references are effectively one level deep with easy navigation — matches the anchor, not 4, since organization and signaling are consistently strong.

5 / 5

Total

19

/

20

Passed

Description

76%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.

A strong description with an explicit what and multiple concrete when-clauses, plus effective sibling-skill disambiguation. Its main weakness is trigger vocabulary that mixes natural phrasing with project-internal jargon ('external-adoption question', 'oracle panel') and misses common synonyms a user would naturally say when seeking a judgment or opinion.

Suggestions

Add natural-language synonyms to the when-clauses, e.g. 'Use when the user asks for an opinion, verdict, or recommendation on adopting a tool, library, or approach'.

Translate project jargon ('oracle panel', 'external-adoption question') into user-facing phrasing, keeping the jargon terms only as secondary keywords.

Mention the deliverable forms (graded verdict, holistic take, position) so the what-clause conveys the full range of outcomes.

DimensionReasoningScore

Specificity

Names the core action ('Judge a supplied subject against the project's evidence and constraints') and enumerates several concrete subject shapes ('an external-adoption question, a holistic take on a document, or a supplied approach set') plus panel capabilities ('consult other models and reconcile their opinions'). Not 5 because it omits the deliverable forms (graded verdict, position) that would make coverage comprehensive; not 3 because it lists well more than 1-2 concrete actions.

4 / 5

Completeness

Explicitly answers what ('Judge a supplied subject against the project's evidence and constraints') and when with concrete trigger clauses ('Use when assessing an external-adoption question, a holistic take on a document, or a supplied approach set. Use for an oracle panel...'), plus sibling-skill routing. Clearly matches the anchor for explicit what AND when with concrete trigger phrases, not 4 since the 'when' is already fully explicit and specific.

5 / 5

Trigger Term Quality

Includes some natural phrases ('a holistic take on a document', 'assessing', 'reconcile their opinions') but leans on project jargon ('external-adoption question', 'oracle panel') and misses common variations a user would naturally say ('give an opinion', 'should we adopt X', 'point of view', 'review these options'). Falls between 'some relevant keywords' (3) and 'good keyword coverage' (4), noticeably closer to 3 because the jargon terms are not what users organically say.

3 / 5

Distinctiveness Conflict Risk

Carves a distinct niche (grounded verdict formation against project evidence) and explicitly disambiguates sibling skills ('Use ce-explain for understanding and ce-doc-review for findings review'). Not 5 because generic verbs like 'judge' and 'take on a document' leave minor overlap risk with general review/opinion skills; not 3 because the explicit sibling routing and specific triggers go beyond 'somewhat specific'.

4 / 5

Total

16

/

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

frontmatter_unknown_keys

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

Warning

Total

15

/

16

Passed

Repository

EveryInc/compound-engineering-plugin

Path

skills/ce-pov/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.