improve-loop

Systematically review and improve every shell feature and builtin command. Iterates through each feature/command, runs code-review, fixes issues, and re-reviews until clean.

1.56x

Quality

68%

Does it follow best practices?

Impact

91%

1.56x

Average score across 3 eval scenarios

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./.claude/skills/improve-loop/SKILL.md

Quality

Content

70%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 thorough, highly actionable improve-loop workflow with excellent sequencing and validation checkpoints, held back by verbosity in its enforcement tone and a monolithic single-file structure that doesn't leverage progressive disclosure.

Suggestions

Move the detailed review dimensions (A-K) and per-agent review instructions into a separate reference file (e.g., REVIEW_DIMENSIONS.md) and link to it from the main loop, reducing SKILL.md to the orchestration overview.

Trim the emphatic 'STOP — READ THIS' and repeated gate-check prose to the essential rule; the TaskList gating already enforces ordering.

Replace the '[paste dimensions below]' placeholder in the agent-launch example with a concrete instruction to read the extracted reference file.

Dimension	Reasoning	Score
Conciseness	Dense with project-specific actionable detail (RULES.md compliance, callCtx.OpenFile, sandbox wrapper) rather than generic concepts, but the emphatic 'STOP — READ THIS' framing and repeated security/gate-check language could be tightened.	3 / 5
Actionability	Provides concrete copy-paste bash commands (gh pr view, ls -d, find, go test), explicit commit-message formats, and structured output templates, with only minor placeholders like '[paste dimensions below]'.	4 / 5
Workflow Clarity	Clear sequence (Step 1 -> 2 loop -> 3 -> 4) with gate checks, explicit validation (2D tests with a 3-attempt retry cap, 3A-3E full sweep), and fix->retest feedback loops for this batch/destructive workflow.	5 / 5
Progressive Disclosure	Well-sectioned with clear headers, but it is a single 551-line monolith with no bundle files; the detailed review dimensions A-K are inlined content that would benefit from being split into a separate reference file.	3 / 5
	Total	15 / 20 Passed

Description

66%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 clear, specific description of a review-and-improve loop for shell features/builtins with good trigger terms and a distinct niche, weakened by the absence of any explicit "Use when..." trigger guidance.

Suggestions

Add an explicit 'Use when...' clause stating the natural trigger phrases (e.g., 'Use when the user asks to review or harden shell builtins/features, or mentions improving command coverage').

Broaden trigger terms with synonyms users might say (e.g., 'harden', 'audit', 'shell commands', 'builtins').

Mention concrete outcomes in the what (e.g., adding tests, fixing bugs, enforcing sandbox) to push specificity toward comprehensive.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ("review and improve", "Iterates through each feature/command", "runs code-review", "fixes issues", "re-reviews until clean") but stops short of comprehensive coverage (no testing/committing).	4 / 5
Completeness	The "what" is clearly stated but there is no "Use when..." clause or equivalent explicit trigger guidance, so completeness is capped at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	Includes natural terms a user would say ("review", "improve", "shell feature", "builtin command", "code-review") with good coverage, but lacks synonyms or file extensions to reach comprehensive.	4 / 5
Distinctiveness Conflict Risk	Targets a clear niche ("every shell feature and builtin command") that is mostly distinct from generic code-review skills, with only minor overlap risk.	4 / 5
	Total	15 / 20 Passed

Validation

87%

Warnings & errors only

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

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (552 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: DataDog/rshell
Path: .claude/skills/improve-loop/SKILL.md
Commit: 30a7e9e

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.