Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/amitrockbear-create-pr
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/create-pr
What's inside
- Skill name
create-pr- Description
- Use when the user asks to file, open, or create a PR.
- License
- MIT
- Instructions
- 76 lines in SKILL.md · view on GitHub
Files (2)
- SKILL.md
- openai.yaml · agents/
SKILL.md (first lines)
# Create PR
Before filing, check whether a PR for this branch already exists. Review the diff locally against the intended PR base (the parent branch for a stack, normally `origin/main` otherwise) to make sure its contents match the goal.
This skill owns PR titles, descriptions and publication for pstack's [Opening a PR](../poteto-mode/playbooks/opening-a-pr.md) workflow. Use existing [verify-change](../verify-change/SKILL.md) evidence; UI checks require the user's explicit request. Write the body and supporting artifacts under a task directory in `~/.pstack/outputs`.
## PR title
- If a Linear ticket ID was mentioned in the conversation, prefix the PR title with `<ticket_id>:` followed by a space (for example, `ABC-1234:`).
- Prefer a concise, human-readable title that explains why the change matters.
BAD
> ❌ negotiate permessage-deflate on the websocket
GOOD
> ✅ ABC-1234: cut websocket frame size by 70%+ with gzipping
## PR description
- Open the description with a simple explanation of the problem based on the user's original prompt, then briefly explain the solution. Do not lead with an implementation inventory. Then, have in bold `Changes:` and then a list of bullet points that describe the changes made in the PR on a high level (dont mention the files or lines that were changed, nor that tests were added or modified).
- After Changes, add **Tradeoffs:** for rejected alternatives a reviewer would reasonably ask about, with a brief reason for the choice. Omit it when there was no meaningful choice.
- Then add **Blast radius:** in one to three sentences explaining who or what is affected and why the change is safe or risky. Omit it when it adds nothing beyond the opening explanation.
- Add **Validation:** with brief live UI/API checks and outcomes, followed by `**Before:**` / `**After:**` evidence: uploaded images/videos, or a short fenced API/server log proving the behavior. Use `None` for missing checks or evidence. Tests/lint do not count as evidence.
- Add a blurb to the end of the PR description about what model is making the changes.
- Use bold labels for PR description sections, including optional sections; keep Markdown headings for skill documentation only.
### Example
Save the description as `body.md` under the task output directory. Use matching absolute media paths in the body and `--attach`; `gh` replaces them with hosted URLs. Verify the published body contains hosted evidence URLs, or report the upload failure.
````md
Refreshing a thread hid its running indicator. The list now preserves live status after refresh.
**Changes:**
- Include the current running state in thread summaries
**Tradeoffs:**
Use the server's running state rather than infer it from cached messages, which can be stale after refresh.
**Blast radius:**
Thread summaries and the list indicator are affected. Message content and task execution are unchanged.
**Validation:**
- UI: refreshed a running thread; indicator stayed visible.
- API: fetched thread summaries; running state was present.
**Before:** None
**After:**
[Attach the reviewed video here.]
```text
GET /threads → 200
{"id":"thread-123","sidebarIndicator":"running"}
```
Made with <model>.
````
### Create/Edit a PR
From the repository checkout, create a PR using the task's absolute body and attachment paths:
```bash
gh pr create --title "Preserve running status after refresh" --base "<base>" --head "<branch>" --body-file "~/.pstack/outputs/<task>/body.md" --attach "~/.pstack/outputs/<task>/refresh.mp4"
gh pr view "<PR URL>" --json body --jq .body
```
For an existing PR, use `gh pr edit "<PR URL>" --body-file "<task-output-dir>/body.md" --attach "<task-output-dir>/refresh.mp4"` with its updated body and absolute paths.
Open a real PR rather than a draft so review bots run. In T3, register existing PRs when starting work and new PRs immediately after creation with `link_pull_request` using the full URL. If the user also asked to babysit it, continue with the bundled [babysit](../babysit/SKILL.md) policy and the workflow's stack timing.From AmitRockBear/pstack-t3code-claudex-port, MIT.
More from AmitRockBear/pstack-t3code-claudex-port
- 0.565score
html-communication
AmitRockBear
Use for plans, specs, write-ups, findings, summaries, reports or comparisons presented as readable HTML; use for UI mocks when the user explicitly asks to…
SkillAFreshness: AClaude Code +9 - 0.561score
correct
AmitRockBear
On explicit /correct, find repeated repository mistakes and prevent them through architecture, types, lint or behavioral tests.
SkillAFreshness: AClaude Code +9 - 0.544score
create-verification-skill
AmitRockBear
Generate a project-local verification skill that drives your app the way a user does — any language, framework, or platform.
SkillAFreshness: AClaude Code +9 - 0.541score
interrogate
AmitRockBear
Use for "interrogate", "adversarial review", "multi-model review", "challenge this", "stress test this code", "find blind spots", or "tear this apart".
SkillAFreshness: AClaude Code +9 - 0.541score
how
AmitRockBear
Use for "how does X work", code walkthroughs before changing something, and placement / ownership / layering questions ("where should this live", "which…
SkillAFreshness: AClaude Code +9 - 0.541score
architect
AmitRockBear
Sketch types, signatures, and module structure before code, then stay in the loop while implementation fills in.
SkillAFreshness: AClaude Code +9 - 0.539score
benchmark-checklist
AmitRockBear
Vet a perf measurement (limiter, tuning, limits, errors, repeatability, relevance, and whether the work happened) before you report or act on it.
SkillAFreshness: AClaude Code +9 - 0.538score
babysit
AmitRockBear
Use when the user asks to monitor, watch, subscribe or babysit a PR.
SkillAFreshness: AClaude Code +9 - 0.536score
arena
AmitRockBear
Spawn N parallel candidates at the same task, pick a base, graft the strongest parts of the losers into it.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 23, 2026 (6 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
Same skill elsewhere
Copies of this SKILL.md published by others; search shows only one of them.
More from AmitRockBear
- 0.569score
use-codex
AmitRockBear
Use when the user asks to use Codex, GPT, Sol, Luna or Astra.
SkillAFreshness: AClaude Code +9 - 0.566score
typescript-best-practices
AmitRockBear
Use when reading or editing any .ts or .tsx file.
SkillAFreshness: AClaude Code +9 - 0.565score
verify-change
AmitRockBear
Use when verifying an implemented change or PR before handoff.
SkillAFreshness: AClaude Code +9 - 0.565score
html-communication
AmitRockBear
Use for plans, specs, write-ups, findings, summaries, reports or comparisons presented as readable HTML; use for UI mocks when the user explicitly asks to…
SkillAFreshness: AClaude Code +9 - 0.565score
use-cursor
AmitRockBear
Use when the user asks to run a model fable-5-1 or opus-5-5 through Cursor, Grok, including follow-ups to an existing Cursor task.
SkillAFreshness: AClaude Code +9 - 0.564score
use-claude
AmitRockBear
Use when the user asks Claude, Opus or Fable for a task, including follow-ups to an existing Claude task.
SkillAFreshness: AClaude Code +9
Cite this page
create-pr skill. Agent Mart. https://agentmart.dev/skills/amitrockbear-create-pr (last updated 2026-10-24)