Skip to content
Back to skills

Embed

ASecurity

Address a PR's review findings — dispatch the fixer to change the code and commit on the branch

  • 67 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 23, 2026
ai-agents

Security analysis

A100/100

Pro scans all 21 files and shows the line behind each finding

Scanned October 2, 2026

npx -y skills add gethuman-sh/human --skill embed --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Embed?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Embed
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/gethuman-sh-embed-d42c904b/badge)](https://www.skillsdirectory.com/skills/gethuman-sh-embed-d42c904b)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: human-pr-fix
description: Address a PR's review findings — dispatch the fixer to change the code and commit on the branch
argument-hint: <key> --pr=<number> --branch=<branch>
---

<!-- human:include launch-briefing -->

`$ARGUMENTS` is `<KEY> --pr=<number> --branch=<branch>` — the PM ticket key, the open pull request whose review findings need addressing, and its branch. Parse them, then delegate to the **human-pr-fixer** agent.

Run the fixer at the `sonnet` tier: implementing a fix is visible-failure work — a red re-review or a failed check catches what it misses — so the expensive tier is not warranted here.

```
Task(subagent_type="human-pr-fixer", model="sonnet", prompt="Address review comments on PR <number> for ticket <KEY> --branch=<branch>", run_in_background=false)
```

The agent reads the machine reviewer's findings from `stage.pr-review` (**and any comment a human left on the PR out of band**), addresses them, commits on the local branch — it does not push; the daemon ships the branch at merge — and records its exit (`done | needs-input | outage`) plus the post-fix head in `stage.pr-fix`. The daemon re-reviews the new commit after a `done`, escalates on `needs-input`, and parks the card uncharged on an `outage`, so you do **not** post board markers or decide the next step yourself: run the agent and report what it changed.

Files in this skill

  • brainstorm-codebase-agent.md3.3 KB
  • brainstorm-opportunities-agent.md3.7 KB
  • brainstorm-recon-agent.md3.3 KB
  • brainstorm-trajectory-agent.md3.4 KB
  • brainstorm-triage-agent.md4.5 KB
  • codenav-skill.md2.6 KB
  • features-recon-agent.md4.8 KB
  • features-synthesis-agent.md10.2 KB
  • findbugs-api-agent.md5.8 KB
  • findbugs-concurrency-agent.md5.6 KB
  • findbugs-errors-agent.md5 KB
  • findbugs-logic-agent.md4.8 KB
  • findbugs-recon-agent.md4.6 KB
  • findbugs-triage-agent.md4.7 KB
  • gardening-complexity-agent.md5.8 KB
  • gardening-duplication-agent.md5.2 KB
  • gardening-hygiene-agent.md6.6 KB
  • gardening-structure-agent.md5.3 KB
  • gardening-survey-agent.md7.1 KB
  • gardening-triage-agent.md8.6 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.