CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

arn-code-batch-implement

This skill should be used when the user says "batch implement", "implement all", "batch execution", "implement all features", "parallel implement", "implement in parallel", "arness batch implement", "arn-code-batch-implement", "run batch implementation", "implement everything", "launch batch workers", or wants to spawn parallel worktree-isolated background agents to implement multiple pending features simultaneously. Each worker runs as a full independent session with all tools. This skill requires pending plans in .arness/plans/ — run arn-code-batch-planning first if none exist.

Quality

86%

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

81%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 well-engineered orchestration skill body: fully executable commands, an explicitly sequenced workflow with validation checkpoints and bounded retry loops appropriate to batch operations, and real one-level-deep reference files for the preflight procedure and worker instructions. Its weaknesses are token efficiency (rationale prose, worked examples, redundancy across worktree steps) and a body length that keeps it from being a lean overview over its references.

Suggestions

Trim rationale prose and the Step 6 worked example (~10 lines): keep the rules ('prune stale metadata', 'do not retry more than once') and drop the justifications and illustrative walkthrough — the deterministic procedure is already unambiguous.

Move the Step 4a collision-resolution case analysis (path/branch collision, already-provisioned reuse, -<n> incrementing) into a references/worktree-provisioning.md, keeping only the create command, the ready/deferred split, and the empty-slug fallback inline in SKILL.md.

Compress the Step 6.5 sample findings table to one example row (or move it into the aggregation reference) and state the column schema in a single line, cutting roughly half of that step's token cost.

DimensionReasoningScore

Conciseness

The body is dense and mostly efficient, but includes rationale prose that could be trimmed ('the single most common cause of transient worktree add failures', 'avoids racing with live workers on .git/worktrees/ metadata and keeps the orchestrator's output easy to read') and a ~10-line worked example, plus some redundancy between Step 4a/4b and the Step 6 retry procedure. Mostly efficient with some unnecessary explanation, matching anchor 3 rather than 4.

3 / 5

Actionability

Fully executable guidance throughout: copy-paste-ready commands ('git -C "$REPO" worktree add "$REPO/.claude/worktrees/arn-batch-<slug>" -b "arn-batch/<slug>"', 'git -C "$REPO" worktree prune'), exact AskUserQuestion prompts with numbered options, exact status strings ('failed -- worktree provisioning (retry exhausted)'), and fully specified Agent spawn parameters (no isolation, run_in_background: true, 5-worker cap). Covers the common cases per anchor 5.

5 / 5

Workflow Clarity

A batch operation with a clearly sequenced Steps 0-7 workflow, explicit validation checkpoints (pre-flight validation, verifying each worktree appears in 'git worktree list', per-feature exit-code/stderr capture, ready/deferred split), and error-recovery feedback loops (prune -> retry once -> cap, spawn-error deferral, findings aggregation with zero-fallback for missing lines). Matches anchor 5 including the batch-operation feedback-loop requirement.

5 / 5

Progressive Disclosure

Both referenced bundle files exist (references/preflight-validation.md, references/worker-instructions.md), are one level deep, and are clearly signaled with full ${CLAUDE_PLUGIN_ROOT} paths; the body is well-sectioned with headers. Not a 5 because the 246-line body inlines procedural detail that could live in a reference (Step 4a collision case analysis, the Step 6 worked example, Step 6.5 findings-table example), making it a full workflow spec rather than a lean overview.

4 / 5

Total

17

/

20

Passed

Description

86%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 with explicit 'Use when' trigger guidance and an extensive list of natural trigger phrases including exact skill-name invocations. The what/when/prerequisite structure is explicit and in third person. Its only weakness is that it describes the mechanism (spawning workers) rather than enumerating the orchestrator's concrete actions.

DimensionReasoningScore

Specificity

Concrete but narrow: 'spawn parallel worktree-isolated background agents to implement multiple pending features simultaneously' and 'Each worker runs as a full independent session with all tools' name the domain with 1-2 concrete actions, but the description omits the skill's own orchestration behaviors (pre-flight validation, progress tracking, handoff). Matches anchor 3 rather than 4 because it does not list several distinct specific actions.

3 / 5

Completeness

Explicitly answers both questions in third person: 'what' via 'spawn parallel worktree-isolated background agents to implement multiple pending features simultaneously. Each worker runs as a full independent session with all tools', and 'when' via 'This skill should be used when the user says...' followed by concrete trigger phrases, plus a prerequisite ('run arn-code-batch-planning first if none exist').

5 / 5

Trigger Term Quality

Comprehensive natural trigger coverage including synonyms and exact command names: 'batch implement', 'implement all', 'parallel implement', 'implement in parallel', 'arness batch implement', 'arn-code-batch-implement', 'launch batch workers', 'implement everything'. This matches the anchor-5 example of comprehensive coverage including synonyms.

5 / 5

Distinctiveness Conflict Risk

Clear batch-orchestration niche with distinct triggers ('batch', 'parallel', 'launch batch workers'), but broad phrases like 'implement all' / 'implement everything' create minor overlap risk with single-feature implementation skills. Mostly distinct with minor overlap, matching anchor 4 rather than 5.

4 / 5

Total

17

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

AppsVortex/arness

Path

plugins/arn-code/skills/arn-code-batch-implement/SKILL.md

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.