agentic-jujutsu

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

2.62x

Quality

49%

Does it follow best practices?

Impact

84%

2.62x

Average score across 3 eval scenarios

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

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

Quality

Content

50%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 highly actionable — concrete executable code, complete API tables, and explicit validation rules — but it is roughly 2-3x longer than needed due to pervasive duplication of the same trajectory pattern, and it reads as product documentation rather than a skill workflow. No sequenced process with validation checkpoints exists, and all reference material is inlined with dangling documentation links.

Suggestions

Collapse the duplicated trajectory examples into one canonical pattern in Quick Start and cut the Advanced Use Cases, Examples, and Best Practices sections that restate it, targeting roughly a third of current length.

Replace the capability catalog with one sequenced workflow (install → initialize wrapper → track trajectory → finalize with honest score → query suggestions) with explicit validation checkpoints between steps.

Move the full API reference tables and use-case examples into reference files under references/ and link them one level deep, removing or fixing the dangling docs/... links.

Dimension	Reasoning	Score
Conciseness	The ~645-line body re-demonstrates the same startTrajectory/addToTrajectory/finalizeTrajectory pattern in Quick Start, Core Capabilities, four Advanced Use Cases, Best Practices, and two Examples, plus marketing padding ("23x faster than Git", "Lock waiting: 0 min (∞)", "Status: ✅ Production Ready"). This is noticeably verbose with several padded/duplicated sections (anchor 2), not the mostly-efficient profile of 3.	2 / 5
Actionability	Mostly executable JavaScript against a concrete API (JjWrapper methods, getSuggestion, getPatterns) with full API tables, explicit validation limits (10KB task cap, 0.0-1.0 scores), and troubleshooting snippets with concrete thresholds. Minor gaps keep it below 5: undefined placeholders (executeOperation, verifyDeployment, executeTask, agent.analyze) and a Best Practices snippet that finalizes a trajectory without starting one.	4 / 5
Workflow Clarity	The body is a capability catalog rather than a sequenced workflow — Quick Start gives install and basic usage, but there is no ordered process with validation checkpoints. Error-handling feedback loops appear only scattered (Use Case 3 try/catch, validation-error troubleshooting), so sequence is present but checkpoints are missing or implicit (anchor 3), not the mostly-checkpointed profile of 4.	3 / 5
Progressive Disclosure	Section structure is clear with good headers, but ~660 lines inline everything — the full API reference, four use cases, and two example sections — that belongs in separate reference files, and the Related Documentation links (docs/AGENTDB_GUIDE.md, docs/VALIDATION_FIXES_v2.3.1.md, package README) do not resolve to any bundle files. This matches anchor 3 (structure present, references not clearly signaled, separable content inline); it is not 2 because organization is far from minimal.	3 / 5
	Total	12 / 20 Passed

Description

48%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 names a distinct niche (agent-oriented version control) but relies on buzzwords instead of concrete actions and provides no "Use when..." trigger guidance. It answers "what" reasonably but leaves "when" entirely to inference, weakening both completeness and trigger-term quality.

Suggestions

Add an explicit 'Use when...' clause naming concrete triggers, e.g. 'Use when multiple AI agents need to commit, branch, or merge concurrently without lock conflicts, or when git lock contention blocks parallel agents.'

Replace buzzword phrases ('Quantum-resistant', 'ReasoningBank intelligence', 'self-learning') with 2-3 concrete capabilities users can evaluate, such as 'tracks operations, learns from successful trajectories, and suggests next steps'.

Include natural trigger terms users would actually say — git, commit, merge, branch, conflicts, concurrent agents, jj — to improve keyword coverage and distinctiveness.

Dimension	Reasoning	Score
Specificity	The description names the domain ("version control for AI agents") but lists no concrete actions — "self-learning", "ReasoningBank intelligence", "multi-agent coordination" are generic buzzwords, and "Quantum-resistant" is an over-claim. This matches anchor 2 (domain named, actions minimal or generic) rather than 3, which requires at least 1-2 concrete actions.	2 / 5
Completeness	The "what" is present (self-learning, multi-agent version control) but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the judging guidelines. It is not 4 because the "when" is entirely absent rather than merely imprecise.	3 / 5
Trigger Term Quality	"version control", "AI agents", and "multi-agent coordination" are relevant keywords, but the natural phrases a user would actually say — git, commit, merge, conflicts, concurrent edits, jj — are missing. Some relevant keywords without common variations matches anchor 3; it is not 4 because the everyday VCS vocabulary users naturally use is largely absent.	3 / 5
Distinctiveness Conflict Risk	"Version control for AI agents with multi-agent coordination" carves a mostly distinct niche with minor overlap risk against generic Git/version-control skills. It is not 5 because no concrete trigger phrases establish a fully clear niche with minimal conflict risk.	4 / 5
	Total	12 / 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 (663 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/RuVector
Path: .claude/skills/agentic-jujutsu/SKILL.md
Commit: 4b59352

Reviewed: 3 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.