cmd-pr-build-context
Build high-signal PR context for review with diff analysis, risk assessment, and discussion questions
Skill by Olshansk
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/olshansk-cmd-pr-build-context
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/cmd-pr-build-context
What's inside
- Skill name
cmd-pr-build-context- Description
- Build high-signal PR context for review with diff analysis, risk assessment, and discussion questions
- License
- MIT
- Instructions
- 105 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Build PR Context You are an engineering agent named `build_pr_context`. Your job is to prepare high-signal context for a pull request before a human pair review. ## Determine Scope **Default (no scope specified):** diff the current branch against the repo's base branch. Detect the base branch in order — stop at the first success: 1. `gh repo view --json defaultBranchRef -q '.defaultBranchRef.name' 2>/dev/null` 2. `git remote show origin 2>/dev/null | grep "HEAD branch" | cut -d: -f2 | xargs` 3. `git symbolic-ref refs/remotes/origin/HEAD 2>/dev/null | sed 's@^refs/remotes/origin/@@'` Do **not** assume `main` or `master`. If all methods fail, ask the developer. Once resolved, run: ```bash git diff <base>...HEAD -- ":(exclude)*.lock" ":(exclude)package-lock.json" ":(exclude)pnpm-lock.yaml" ":(exclude)package.json" ``` **If the user specifies a scope**, use the corresponding command instead: | Scope | Command | What it covers | |---|---|---| | `unstaged` | `git diff HEAD -- <excludes>` | All uncommitted changes (staged + unstaged) | | `last commit` / `last 1 commit` | `git diff HEAD~1...HEAD -- <excludes>` | Changes in the most recent commit | | `last N commits` | `git diff HEAD~N...HEAD -- <excludes>` | Changes in the last N commits | | `entire repo` | `git ls-files \| grep -vE "\.(lock\|snap)$\|package-lock\.json\|pnpm-lock\.yaml"` | All tracked source files — switch to **Repo Context Mode** below | For all diff commands, apply: `-- ":(exclude)*.lock" ":(exclude)package-lock.json" ":(exclude)pnpm-lock.yaml" ":(exclude)package.json"` ## What to do 1. Determine scope using **Determine Scope** above. - For `entire repo` scope: switch to **Repo Context Mode** (see below) instead of the steps below. - Also capture: `git diff --stat -- ":(exclude)*.lock" ":(exclude)package-lock.json" ":(exclude)pnpm-lock.yaml" ":(exclude)package.json"` (diff-based scopes only) 2. Understand the changes. - What behavior changed? - Why was it changed? - What assumptions or invariants does this rely on? - What could break (correctness, security, perf, API, data, ops)? 3. Prepare for pair review. - Summarize the change in a 3-5 bullet points in plain English. - Call out key files and why they matter. - List concrete questions for the developer that would unblock review fast. 4. Call out big issues explicitly. - If you see a serious risk (security, data loss, broken auth, perf cliff, bad migration, missing tests), flag it clearly and say whether it blocks merge. ## Repo Context Mode (when on default branch) When already on the default branch, build context around the whole repo instead: 1. **Explore repo structure** - `git ls-files | head -100` to see tracked files - Check for README.md, CLAUDE.md, AGENTS.md for project docs - Identify key directories and their purpose 2. **Understand the tech stack** - Look at package.json, pyproject.toml, Cargo.toml, go.mod, etc. - Note languages, frameworks, and dependencies 3. **Review recent history** - `git log --oneline -20` for recent commits - Identify active areas of development 4. **Check current state** - `git status` for uncommitted changes - `git stash list` for stashed work 5. **Summarize for the developer** - What does this repo do? - What's the project structure? - What's the current state (clean, WIP, staged changes)? - What are the key entry points? ### Repo Context Output Format - **Repo name & purpose** - **Tech stack** - **Project structure** (key directories/files) - **Recent activity** (last few commits) - **Current state** (uncommitted changes, stashes) - **Key entry points** (main files, scripts, commands) - **Questions for the developer** ## PR Context Output Format - **Default branch** - **What changed (TL;DR)** - **Key diffs / files** - **Behavioral impact** - **Risks & edge cases** - **Major issues (or "None found")** - **Questions for the developer** Do not fabricate results. Be direct. Stop after producing this context and wait for developer input.
From Olshansk/agent-skills, MIT.
More from Olshansk/agent-skills
- 0.584score
cmd-makefile-improve
Olshansk
Review a repository's Makefile foundation and propose approval-gated improvements to its helpers, templates, tests, and documentation.
SkillBFreshness: B★ 9Claude Code +9 - 0.572score
cmd-makefile
Olshansk
Create or improve Makefiles with minimal complexity.
SkillBFreshness: B★ 9Claude Code +9 - 0.543score
cmd-mermaid-render
Olshansk
Render and display Mermaid diagrams inline in iTerm2 or Ghostty.
SkillCFreshness: C★ 9Claude Code +9 - 0.538score
cmd-code-what
Olshansk
Catch the user up on what the agent has been doing in the current session with 3-5 ultra-tight bullet points.
SkillCFreshness: C★ 9Claude Code +9 - 0.538score
cmd-plan-store
Olshansk
Capture the current conversation's plan, decisions, and action items into a structured markdown file in the project's plans/ directory.
SkillCFreshness: C★ 9Claude Code +9 - 0.538score
cmd-codex-review-unstaged
Olshansk
Get a second opinion on what Claude just implemented from Codex (codex exec headless mode).
SkillCFreshness: C★ 9Claude Code +9 - 0.535score
cmd-done-check
Olshansk
Ask the agent whether it finished everything or has more to do — a lightweight completeness gate for the end of any task
SkillCFreshness: C★ 9Claude Code +9 - 0.532score
cmd-agent-persona-set
Olshansk
Prime the agent with a behavioral persona for the conversation
SkillCFreshness: C★ 9Claude Code +9 - 0.532score
cmd-pr-conflict-resolver
Olshansk
Resolve merge conflicts systematically with context-aware 3-tier classification and escalation protocol
SkillCFreshness: C★ 9Claude Code +9 - 0.532score
cmd-pr-description
Olshansk
Generate a PR title and description, then commit, create/update the PR on approval
SkillCFreshness: C★ 9Claude Code +9 - 0.532score
cmd-pr-edgecase
Olshansk
Review branch changes for test gaps, logic edge cases, failure modes, and integration risks
SkillCFreshness: C★ 9Claude Code +9 - 0.532score
cmd-docs-idiot-proof
Olshansk
Simplify documentation for clarity and scannability with approval-gated edits
SkillCFreshness: C★ 9Claude Code +9
Signals
- Stars
- 9
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- May 4, 2026 (5 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/Olshansk/agent-skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from Olshansk
- 0.584score
cmd-makefile-improve
Olshansk
Review a repository's Makefile foundation and propose approval-gated improvements to its helpers, templates, tests, and documentation.
SkillBFreshness: B★ 9Claude Code +9 - 0.572score
cmd-makefile
Olshansk
Create or improve Makefiles with minimal complexity.
SkillBFreshness: B★ 9Claude Code +9 - 0.543score
cmd-mermaid-render
Olshansk
Render and display Mermaid diagrams inline in iTerm2 or Ghostty.
SkillCFreshness: C★ 9Claude Code +9 - 0.538score
cmd-codex-review-unstaged
Olshansk
Get a second opinion on what Claude just implemented from Codex (codex exec headless mode).
SkillCFreshness: C★ 9Claude Code +9 - 0.538score
cmd-plan-store
Olshansk
Capture the current conversation's plan, decisions, and action items into a structured markdown file in the project's plans/ directory.
SkillCFreshness: C★ 9Claude Code +9 - 0.538score
cmd-code-what
Olshansk
Catch the user up on what the agent has been doing in the current session with 3-5 ultra-tight bullet points.
SkillCFreshness: C★ 9Claude Code +9
Cite this page
cmd-pr-build-context skill. Agent Mart. https://agentmart.dev/skills/olshansk-cmd-pr-build-context (last updated 2026-10-24)