CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ce-optimize

Optimize a named target with a measured loop: attribute a workload's cost, or score variants and keep winners. Use when a working system's metric should move and the winning change is not already known. Use ce-debug when the job is diagnosis; use ce-work when the change is already known.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

85%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 tightly written orchestrator for a complex four-phase workflow: it sequences phases with hard gates and explicit validation checkpoints while delegating executable detail to well-signaled, real reference files, and every bundle file is reachable with its purpose stated where it is needed. The only notable improvement is trimming a few redundant discipline statements.

DimensionReasoningScore

Conciseness

The body is dense and purely instructional with no padding about concepts Claude already knows, but a few passages could be trimmed — "A step is done only after it ran. Describing a measurement, dispatch, or checkpoint is not doing it." restates the 'Done when' discipline, and the Execution Surface paragraph packs three definitions into one block — matching anchor 4 rather than the fully lean anchor 5.

4 / 5

Actionability

Concrete specifics appear directly (`git rev-parse --show-toplevel`, `<repo-root>/.compound-engineering/config.yaml`, `.context/compound-engineering/ce-optimize/<spec-name>/`, the `optimize/<spec-name>` branch, and gates like "Do not enter Phase 2 until the user explicitly approves"), but most executable detail (tick protocol, script invocations) is delegated to references, leaving minor gaps in the body itself — anchor 4, not 5's copy-paste-ready coverage.

4 / 5

Workflow Clarity

Four phases run in explicit order, each naming the reference it "cannot start without", with hard gates (clean-tree, user approval, dependency pre-approval), eight enumerated stopping criteria, checkpoints CP-0 through CP-5, a measure-write-verify write order, an explicit error path ("If the run instead stopped at a check it could not pass, say what blocked it"), and a defined resume protocol — full validation and feedback loops per anchor 5.

5 / 5

Progressive Disclosure

The body is a clear overview whose five phase references (persistence.md, spec.md, measurement.md, loop.md, wrap-up.md) all exist as real files and are boldly signaled ("**Read `references/loop.md`**") with stated purpose and phase binding; the remaining bundle files (schemas, example specs, templates, agents, scripts) are each cited with purpose at point of use one level deeper, with no chains that bury content — matching anchor 5.

5 / 5

Total

18

/

20

Passed

Description

71%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, third-person, and cleanly separates what the skill does from when to use it, with explicit disambiguation against sibling skills. Its main weakness is trigger-term coverage: it leans on technical framing ('metric should move', 'attribute a workload's cost') and lacks the everyday synonyms users would naturally say when they need this skill.

Suggestions

Add concrete trigger phrases in the user's voice, e.g. 'Use when the user asks to improve, tune, or speed up a system's measured performance and the fix is not already known.'

Include everyday synonyms such as 'improve', 'tune', or 'make faster' alongside 'optimize' and 'metric' so the description matches how users actually phrase the request.

DimensionReasoningScore

Specificity

"attribute a workload's cost, or score variants and keep winners" lists several concrete actions around the headline 'measured loop', but coverage is not comprehensive (spec construction, harness building, and wrap-up are unnamed), matching anchor 4 rather than 5 while clearly exceeding anchor 3's '1-2 concrete actions'.

4 / 5

Completeness

Both parts are explicit: the 'what' ("attribute a workload's cost, or score variants and keep winners") and an explicit trigger clause ("Use when a working system's metric should move and the winning change is not already known"), plus sibling routing. Not 5 because the when-clause states an abstract condition rather than concrete trigger phrases a user would actually say.

4 / 5

Trigger Term Quality

"Optimize" and "metric" are natural user terms, but common variations such as 'improve', 'tune', 'speed up', or 'performance' are absent, fitting anchor 3 ('some relevant keywords but missing common variations or synonyms') rather than 4, where only 'a few' natural terms would be missing.

3 / 5

Distinctiveness Conflict Risk

"use ce-debug when the job is diagnosis; use ce-work when the change is already known" explicitly routes away from the two sibling skills, and the positive trigger ('metric should move and the winning change is not already known') carves a clear niche with minimal conflict risk, matching anchor 5.

5 / 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-optimize/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.