CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

manage-pull-requests

Manage Harness Code pull requests via MCP. List, create, update, and merge PRs; add reviewers and submit review decisions (approve / request changes); post, update, and delete comments; read the discussion timeline; and check CI status. Use when asked to open a PR, merge a PR, approve a PR, leave a code review comment, request changes, assign reviewers, or inspect PR checks and activity. Trigger phrases: pull request, PR review, merge PR, approve PR, request changes, add reviewer, PR comments, PR checks, PR activity, code review.

75

Quality

94%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

88%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 strong operational reference: fully executable guidance for every PR operation with explicit validation and recovery loops around the destructive merge path. The only weaknesses are repetition of the pr_activity gotcha and the URL shortcut, and a monolithic single-file layout with no progressive split between overview and detail.

Suggestions

State the pr_activity-is-the-only-way-to-read-comments rule once (e.g. in Step 7 plus one bolded callout) and have the tools table, Performance Notes, and Troubleshooting reference it instead of restating it — the same guidance currently appears five times.

Merge the duplicate URL-shortcut guidance: the MCP Tools section and the Performance Notes bullet both explain the url= parameter; keep one and cross-reference it.

Consider splitting the Troubleshooting section (or extended merge/checks detail) into a reference file so SKILL.md stays a lean overview — there are currently no bundle files and all ~285 lines are inlined.

DimensionReasoningScore

Conciseness

Every section is operational (concrete parameter blocks, zero concept-explanation padding), but the pr_activity/POST-only gotcha is repeated five times (tools table, Step 7, Performance Notes, and two Troubleshooting entries) and the URL shortcut is stated in both the MCP Tools section and Performance Notes — minor over-explanation that could be trimmed.

4 / 5

Actionability

Every operation ships a copy-paste-ready parameter block with concrete values (e.g. 'path: "scheduler/queue.go", line_new: 142', 'decision: "approved" | "changereq"'), and exact accepted values plus the harness_schema tool for body fields are documented — fully executable coverage of the common cases.

5 / 5

Workflow Clarity

Steps 1–10 are clearly sequenced, with explicit validation before the destructive merge ('Set dry_run: true first to preview the merge', 'Use this before merging to confirm required checks are green') and explicit error-recovery loops (drop or re-fetch source_sha on SHA mismatch). The destructive-operation cap does not apply because validation is present.

5 / 5

Progressive Disclosure

Well-organized single file (tools table, numbered steps, examples, performance notes, troubleshooting) that delegates exact JSON schemas to the harness_schema tool instead of inlining them, but at ~285 lines all content is inline with no split between quick-start and deep reference, so it fits anchor 4 rather than anchor 5's 'content appropriately split'.

4 / 5

Total

18

/

20

Passed

Description

96%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.

An exemplary description: third-person voice, comprehensive concrete actions, explicit 'Use when' triggers with natural phrasings and synonyms. The only weakness is a couple of generic trigger phrases ('code review') that carry minor conflict risk with adjacent skills.

DimensionReasoningScore

Specificity

'List, create, update, and merge PRs; add reviewers and submit review decisions (approve / request changes); post, update, and delete comments; read the discussion timeline; and check CI status' enumerates 10+ concrete actions with comprehensive lifecycle coverage, matching anchor 5 rather than anchor 4's 'minor gaps in coverage'.

5 / 5

Completeness

Explicitly answers 'what' (the full action list) and 'when' ('Use when asked to open a PR, merge a PR, approve a PR...') with concrete trigger phrases — an exact match for anchor 5, clearly above anchor 4's 'when could be more explicit'.

5 / 5

Trigger Term Quality

Includes natural user phrasings ('open a PR, merge a PR, approve a PR, leave a code review comment, request changes, assign reviewers') plus an explicit trigger list with synonyms ('pull request' and 'PR', 'PR review', 'PR comments', 'PR checks') — comprehensive natural-term coverage per anchor 5.

5 / 5

Distinctiveness Conflict Risk

Scoped to 'Manage Harness Code pull requests via MCP' — a clear niche — but bare trigger phrases like 'code review' and 'PR comments' are generic and could collide with general code-review or git skills, fitting anchor 4 ('minor overlap risk with closely related skills') rather than anchor 5's minimal conflict risk.

4 / 5

Total

19

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

harness/harness-ai

Path

plugins/claude/skills/manage-pull-requests/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.