agentic-jujutsu

Quantum-resistant, self-learning version control for AI agents with ReasoningBank intelligence and multi-agent coordination

1.26x

Quality

50%

Does it follow best practices?

Impact

81%

1.26x

Average score across 3 eval scenarios

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./.agents/skills/agentic-jujutsu/SKILL.md

Quality

Content

50%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with extensive executable code and a clear core workflow, but it is severely over-long and monolithic — everything is inlined in SKILL.md with no progressive disclosure to separate reference files, and validation checkpoints are reactive rather than gated. It reads like product/API documentation rather than a lean skill overview.

Suggestions

Trim the body to a lean overview and move the full API reference, advanced use cases, and examples into separate reference files (e.g., references/API.md, references/EXAMPLES.md) linked one level deep.

Remove redundant multi-agent examples (the same pattern appears in Core Capabilities #4, Best Practices #3, and Examples #2) and drop the marketing 'Performance Characteristics' table.

Add explicit gated validation checkpoints before destructive operations (e.g., 'validate before rebase/merge; only proceed when valid') rather than relying on reactive try/catch.

Dimension	Reasoning	Score
Conciseness	At roughly 640 lines the body is heavily padded: three near-duplicate multi-agent examples, the full API reference repeated as both tables and code, a marketing 'Performance Characteristics' table (23x faster than Git), inline version markers, and a version-history dump — far exceeding a lean SKILL.md overview.	1 / 3
Actionability	Packed with concrete, copy-paste-ready JavaScript (require('agentic-jujutsu'), jj.startTrajectory, await jj.newCommit), an npx install command, and API tables with return types — fully executable guidance, not pseudocode.	3 / 3
Workflow Clarity	The core trajectory sequence (startTrajectory → operations → addToTrajectory → finalizeTrajectory) is clear and there are error-handling feedback loops, but validation is largely input-validation and reactive try/catch rather than gated verify-then-proceed checkpoints before risky ops like rebase, merge, or deploy.	2 / 3
Progressive Disclosure	No bundle files exist and the full API reference, advanced use cases, examples, and best practices are all dumped inline in one monolithic document; the only references point to external package docs rather than well-signaled one-level-deep skill files.	2 / 3
	Total	8 / 12 Passed

Description

50%

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 states a clear domain and several feature areas but is heavy on buzzwords ('Quantum-resistant', 'ReasoningBank intelligence') and omits any explicit 'when to use' trigger guidance. It is moderately specific and distinctive but would benefit from concrete actions and natural trigger terms.

Suggestions

Add an explicit 'Use when...' clause stating when Claude should invoke this skill (e.g., 'Use when multiple AI agents need concurrent, conflict-free version control').

Replace buzzwords ('ReasoningBank intelligence', 'Quantum-resistant') with concrete actions users would recognize (e.g., 'create commits, resolve conflicts, track operations across agents').

Add natural trigger terms like 'commits', 'branches', 'git', or 'multi-agent' that a user would actually say.

Dimension	Reasoning	Score
Specificity	Names the domain ('version control for AI agents') and several capability areas (quantum-resistant security, self-learning, multi-agent coordination), but these are abstract, buzzword-heavy features rather than concrete actions like 'create commits' or 'resolve conflicts'.	2 / 3
Completeness	States what the skill does, but provides no 'Use when...' clause or equivalent trigger guidance for when Claude should invoke it, capping completeness at 2 per the rubric guideline.	2 / 3
Trigger Term Quality	Includes natural terms a user might say ('version control', 'AI agents'), but these are diluted by proprietary jargon ('ReasoningBank intelligence', 'Quantum-resistant') and missing common variations like 'git', 'commits', or 'branches'.	2 / 3
Distinctiveness Conflict Risk	The AI-agent and multi-agent framing carves a niche, but the dominant 'version control' core overlaps with general VCS/git skills and lacks explicit triggers to sharpen the distinction.	2 / 3
	Total	8 / 12 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 (646 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: ruvnet/ruflo
Path: .agents/skills/agentic-jujutsu/SKILL.md
Commit: 26c35b5

Reviewed: about 5 hours 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.