harness-plan

HAR: Research-backed, team-validated task planning, Plans.md management, progress sync. Trigger: create a plan, add tasks, update Plans.md, mark complete, check progress. Do NOT load for: implementation, review, release.

Quality

78%

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 ./opencode/skills/harness-plan/SKILL.md

Quality

Content

73%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, well-structured operational contract with strong workflow clarity and actionable commands. Its weaknesses are repetition that bloats the body and references to scripts that do not ship in the bundle.

Suggestions

Consolidate the repeated Spec delta / Spec skip reason and precedence statements into a single canonical block and reference it, cutting roughly 30-40 lines of duplication.

Move the detailed pre-approval JSON schema field list (plan-preapproval.v2 fields, max_uses semantics) into references/create.md, keeping only the fixed-format template and rules inline.

Clarify which referenced scripts ship to consumers (e.g. note that scripts/plan-registry.sh, plans-issue-bridge.sh, generate-sprint-contract.js are dev-only) or remove them from consumer-facing guidance, as the scripts/ directory is absent from the bundle.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient operational contract text, but repeats the Spec delta / Spec skip reason clause ~4×, states the spec.md > sub-spec > Plans.md precedence twice, and re-defines 'non-trivial' multiple times — these could be consolidated.	3 / 5
Actionability	Concrete commands, file paths, marker tables, and a fixed-format pre-approval template are copy-paste ready, but the 11-step create flow is partly high-level and several referenced scripts do not ship in the bundle.	4 / 5
Workflow Clarity	Multi-step flows are clearly sequenced (11-step create, sync Step 0-6, 9-step quality flow) with explicit validation checkpoints — non-trivial planning gate, plan-time pre-approval, retrospective default-on — and feedback loops for error recovery.	5 / 5
Progressive Disclosure	References are one level deep, clearly signaled with 'See [references/...]', and all four core reference files exist; however the body inlines detailed contract specs (pre-approval JSON schema fields, dual-source rules) and references five scripts that are absent from the bundle.	4 / 5
	Total	16 / 20 Passed

Description

83%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, well-bounded description that clearly states capabilities, triggers, and exclusions in third person. The main weakness is mild fluff adjectives in the 'what' clause and slight overlap with the sibling sync skill.

Suggestions

Drop the 'Research-backed, team-validated' qualifiers from the 'what' clause and lead with concrete verbs (e.g. 'Creates and maintains Plans.md task plans, manages markers, syncs progress with implementation').

Disambiguate the sync trigger from harness-sync, e.g. 'check progress / sync Plans.md state (not implementation sync)'.

Dimension	Reasoning	Score
Specificity	The trigger clause lists five concrete actions ('create a plan, add tasks, update Plans.md, mark complete, check progress'), but the 'what' sentence relies on domain labels padded with mild fluff ('Research-backed, team-validated'), leaving a minor gap versus the fully comprehensive anchor.	4 / 5
Completeness	It explicitly answers both what ('task planning, Plans.md management, progress sync') and when ('Trigger: ...'), plus a concrete negative boundary ('Do NOT load for: implementation, review, release').	5 / 5
Trigger Term Quality	Phrases like 'create a plan', 'add tasks', 'mark complete', and 'check progress' are natural user language with good coverage, though a few common synonyms are missing.	4 / 5
Distinctiveness Conflict Risk	The explicit 'Do NOT load for' carve-out sharply limits conflict, but 'progress sync / check progress' overlaps the sibling harness-sync surface, leaving minor overlap risk.	4 / 5
	Total	17 / 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
relative_links	Relative link issues: 6 missing, 6 deeper-than-1-level	Warning
referenced_paths_exist	Referenced path issues: 8 missing	Warning

	Total	14 / 16 Passed

Repository: Chachamaru127/claude-code-harness
Path: opencode/skills/harness-plan/SKILL.md
Commit: 7d8deba

Reviewed: about 15 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.