aif-evolve

Self-improve AI Factory skills based on project context, accumulated patches, and codebase patterns. Analyzes what went wrong, what works, and enhances skills to prevent future issues. Use when you want to make AI smarter for your project.

Quality

72%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/aif-evolve/SKILL.md

Quality

Content

77%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 an exceptionally actionable, clearly sequenced workflow with strong validation and feedback loops, but it is over-long for a single file: it repeats critical warnings several times and inlines templates/report formats that would benefit from being split into reference files.

Suggestions

Deduplicate the "NEVER edit skills/aif-*" rule — state it once authoritatively and reference it elsewhere instead of repeating the full warning ~5 times.

Move the reusable templates (Prevention Point Registry table, stale-rules report format, evolution-log template, skill-context file template) into ./references/ files and link to them, improving progressive disclosure and trimming the main file.

Condense the repeated explanations of the skill-context-priority-over-base principle into a single stated rule with brief back-references.

Dimension	Reasoning	Score
Conciseness	The workflow is mostly efficient and densely informative, but the "NEVER edit skills/aif-*" rule is restated ~5 times (Critical section, Step 7.2 item 5, Rules #7, Artifact Ownership) and several explanatory paragraphs re-explain the skill-context-priority principle, so it could be tightened; not a 2 because most content is genuinely necessary process detail.	3 / 5
Actionability	Highly executable: exact file paths, glob patterns, a JSON cursor shape, the Prevention Point Registry table format, the skill-context file template, the evolution-log template, and concrete cursor-update rules give copy-paste-ready guidance covering the common cases.	5 / 5
Workflow Clarity	A clear Step 0–Step 8 sequence with explicit validation checkpoints (resolve-and-verify target, "Do NOT proceed to Step 5 until all stale rule decisions collected", enforce outputs against skill-context rules, cursor not advanced on failure) and feedback loops for the batch/destructive skill-file edits.	5 / 5
Progressive Disclosure	Sections are well-organized, but this ~580-line skill is a single monolithic file with no bundle files and inline templates/report-formats (Prevention Point Registry, stale-rules report, evolution log, context-file template) that would be better split into references; the under-50-line exception does not apply, and references are entirely absent rather than clearly signaled.	3 / 5
	Total	16 / 20 Passed

Description

67%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 clearly states what the skill does and includes an explicit 'Use when' trigger in correct third-person voice, but the trigger phrasing is generic and lacks the concrete trigger phrases and synonyms that would make it unambiguous.

Suggestions

Tighten the trigger clause with concrete phrases, e.g. "Use when you want to make AI smarter for your project — e.g. after repeated similar bugs, to turn patches into reusable skill rules, or to evolve installed aif-* skills."

Add natural synonyms users would actually say ("improve skills", "recurring bugs", "learn from patches") to broaden trigger keyword coverage.

Make the concrete actions more comprehensive and specific (e.g. "extracts prevention points from patches, detects stale skill-context rules, writes project-specific rules") to push specificity toward 5.

Dimension	Reasoning	Score
Specificity	Names the domain (AI Factory skill self-improvement) and several concrete actions — "Analyzes what went wrong, what works, and enhances skills to prevent future issues" — using third-person voice; falls short of 5 because the actions stay somewhat abstract rather than enumerating comprehensive concrete capabilities.	4 / 5
Completeness	Explicitly answers both 'what' (self-improve skills from project context, patches, codebase patterns) and 'when' ("Use when you want to make AI smarter for your project"), but the 'when' trigger is generic rather than concrete trigger phrases, so it sits below 5.	4 / 5
Trigger Term Quality	The "Use when you want to make AI smarter for your project" clause is a fairly natural trigger, but keyword coverage is thin — missing common variations like "recurring bugs", "fix repeated mistakes", or "improve skills"; not a 4 because several natural synonyms users would say are absent.	3 / 5
Distinctiveness Conflict Risk	"AI Factory skills", "accumulated patches", and "skill-context" carve a distinct niche with low conflict risk; not a 5 because "make AI smarter" is broad and could mildly overlap with sibling skills like aif-improve/aif-fix.	4 / 5
	Total	15 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (621 lines); consider splitting into references/ and linking	Warning
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	13 / 16 Passed

Repository: lee-to/ai-factory
Path: skills/aif-evolve/SKILL.md
Commit: e144f95

Reviewed: 5 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.