CtrlK
BlogDocsLog inGet started
Tessl Logo

compose-next

Use for multi-step feature work, bug fixes, or refactors where requirements need to settle, a feature document should carry design + tasks + delivery evidence, and the change deserves independent review before merge. Use it only when the user explicitly requests this workflow, whether with `/compose-next`, by name, or in any other clear natural language; do not infer the request from task complexity alone. Not for one-shot edits, single-file tweaks, or answering questions — those need no orchestration overhead.

71

Quality

87%

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

96%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 exceptionally well-crafted orchestration contract: every phase has concrete commands, exact paths, explicit gates, and validation loops, all delivered with no wasted tokens. The only structural improvement available is offloading the spec template and reviewer checklist into a reference file to shorten the always-loaded overview.

DimensionReasoningScore

Conciseness

The ~185-line body is dense imperative instruction with zero concept explanation, zero padding, and no re-teaching of things Claude already knows — e.g. "Never begin implementation on `main` or `master` without explicit user consent", "After two failed fixes, stop patching and re-derive the cause". Every token earns its place, matching the lean-and-efficient anchor.

5 / 5

Actionability

Guidance is fully executable: copy-paste git commands ("git worktree add \"$path\" -b \"$branch\" <base>", "git check-ignore -q \"$path\"", "git rev-parse --short"), an exact spec-document template with frontmatter fields and task syntax, a Report template, and an enumerated reviewer package (workspace path, base/head SHA, diff command, one-line-per-command verification summary). Per the rubric's scoring note, absence of code in an instruction-only skill is not penalized when guidance is this concrete.

5 / 5

Workflow Clarity

The pipeline (Step 0 orient → grill → workspace → spec → implement → verify → review → finalize → finish) is explicitly sequenced with routing rules ("Every path passes through Workspace before Spec or Implement"), explicit validation checkpoints (fresh-evidence verification, PRE-EXISTING baseline marking, verify strictly before review), and a feedback loop for error recovery (fix criticals → re-verify → re-review, with an explicit impasse-report exit when the loop stops converging). Destructive actions are gated ("never auto-approve the destructive option", worktree removal scoped to `.worktrees/`).

5 / 5

Progressive Disclosure

The skill is a single-file contract with no bundle files and no nested references, well-organized under clear phase headings — good structure with easy navigation. It falls short of the 5 anchor because the spec template and the reviewer checklist are inlined in SKILL.md where a one-level-deep reference file could keep the overview leaner, a minor organization gap rather than content that clearly belongs elsewhere.

4 / 5

Total

19

/

20

Passed

Description

78%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: it explicitly states when to invoke, guards against accidental triggering, and scopes out non-matching work. The main weakness is that the "what" describes the workflow's conditions and artifacts rather than the concrete phase pipeline it orchestrates, leaving a reader to infer what actually runs.

Suggestions

Name the concrete phase pipeline (e.g. "Runs grill → worktree → spec → implement → verify → review → finalize") in the description so the "what" states the actions performed, not just their preconditions.

Add a few natural trigger synonyms users might say when requesting the workflow (e.g. "feature spec", "code review before merge", "worktree workflow") to broaden natural-language matching.

State the primary artifact the user ends up with (a feature document plus a reviewed, verified branch) in one concrete clause to sharpen the capability statement.

DimensionReasoningScore

Specificity

The description lists several specific elements — "multi-step feature work, bug fixes, or refactors", "a feature document should carry design + tasks + delivery evidence", and "independent review before merge" — which matches the anchor for several specific actions with minor gaps. It falls short of a 5 because the actual workflow pipeline (grill, worktree, spec, verify, review phases) is never named, and short of vagueness anchors because the artifacts and gates are concrete rather than generic.

4 / 5

Completeness

Both "what" ("a feature document should carry design + tasks + delivery evidence, and the change deserves independent review before merge") and "when" ("Use it only when the user explicitly requests this workflow... whether with "/compose-next", by name, or in any other clear natural language") are explicitly stated, plus negative scope ("Not for one-shot edits, single-file tweaks, or answering questions"). It sits below the 5 anchor because the "what" describes enabling conditions rather than the concrete actions the skill performs.

4 / 5

Trigger Term Quality

Natural phrases users would say are present — "feature work", "bug fixes", "refactors", "review before merge", and the explicit "/compose-next" invocation — giving good keyword coverage. A few natural synonyms and variations (e.g. "PR", "spec doc", "change request") are missing, so it does not reach the comprehensive-coverage anchor.

4 / 5

Distinctiveness Conflict Risk

Invocation is restricted to an explicit user request with explicit anti-drift guidance ("do not infer the request from task complexity alone") and clear exclusions ("Not for one-shot edits, single-file tweaks, or answering questions"), giving it a distinct niche with minimal overlap risk against other workflow or coding skills.

5 / 5

Total

17

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
XiaomiMiMo/MiMo-Code
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.