branch-workspace-lifecycle
Use when starting isolated feature work, deciding whether to use a git worktree, preparing branch cleanup, finishing a development branch, opening a PR…
Skill by fabiocantone
Chalk Labs
Need help deploying branch-workspace-lifecycle?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
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/fabiocantone-branch-workspace-lifecycle
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/branch-workspace-lifecycle
What's inside
- Skill name
branch-workspace-lifecycle- Description
- Use when starting isolated feature work, deciding whether to use a git worktree, preparing branch cleanup, finishing a development branch, opening a PR, merging, or preserving handoff state before integration
- License
- MIT
- Instructions
- 86 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Branch Workspace Lifecycle ## Overview Long-running coding work should not blur workspace state. CoderSteroids uses this skill to decide when work needs an isolated branch or worktree, how to preserve provenance, and how to finish the branch without losing verification or project memory. ## Start Gate Before starting non-trivial implementation: 1. Inspect `git status --short --branch`. 2. Identify whether the current worktree has unrelated changes. 3. Decide whether isolation is needed. 4. Record the branch/workspace choice in the active plan or thread ledger when the task is more than a quick edit. Use an isolated branch or worktree when: - Existing changes are unrelated to the new task. - The task is risky, broad, or likely to take more than one session. - A benchmark, migration, dependency update, or generated-output step may create noisy diffs. - The user asks for parallel work, comparison branches, or a PR-ready result. You may stay in the current workspace when: - The worktree is clean and the task is small. - The user explicitly wants the current branch used. - The repository does not support worktrees cleanly and the fallback is recorded. ## Worktree Decision Prefer the repository's native workflow if it exists. Otherwise use git branches and worktrees conservatively: - Do not overwrite or discard user changes. - Do not delete a worktree or branch without explicit user approval. - Keep worktree names tied to the task, not generic labels. - Record the base branch and target branch. - Record any setup commands needed to reproduce the workspace. If a worktree cannot be created, state the blocker and choose the safest fallback: current branch with explicit status notes, a normal branch, or a separate clone only if the user approves. ## Finishing Gate Before saying a branch is ready to merge, push, or PR: 1. Re-read the active plan, roadmap, and relevant wiki state. 2. Check `git status --short --branch`. 3. Summarize changed files by purpose. 4. Run focused verification and broader affected verification. 5. Run `git diff --check`. 6. Update roadmap/wiki/thread ledger if the work changes direction, decisions, sources, or next action. 7. Decide the finish path with the user when it affects remote state: - leave local only; - commit only; - push branch; - open draft PR; - merge locally; - clean up worktree/branch. Do not treat "tests passed" as sufficient if memory, docs, or handoff artifacts are stale. ## Completion Report When completing branch or worktree lifecycle work, include: ```markdown ## Workspace Lifecycle - Base: - Branch/worktree: - Dirty state before: - Changed files: - Verification: - Memory/docs updated: - Finish decision: - Cleanup needed: ``` ## Common Mistakes - Starting risky work on a dirty branch without naming the risk. - Creating a worktree and then forgetting setup or cache differences. - Merging or deleting branches without explicit user approval. - Pushing changes before verification and memory flush. - Reporting a clean finish while untracked benchmark or generated files remain.
From fabiocantone/codersteroids, MIT.
More from fabiocantone/codersteroids
- 0.474score
prompt-briefing
fabiocantone
Use when turning a short user request into a complete prompt, preparing a Codex/Claude/subagent handoff, rewriting an ambiguous task brief, or asking the user…
SkillCFreshness: CClaude Code +9 - 0.474score
chat-continuity
fabiocantone
Use when starting a new chat, resuming work from another chat, handing off work, or preserving project context across conversations
SkillCFreshness: CClaude Code +9 - 0.474score
code-review-discipline
fabiocantone
Use when requesting code review, reviewing completed work, receiving review feedback, evaluating reviewer suggestions, or deciding whether a change is ready…
SkillCFreshness: CClaude Code +9 - 0.474score
post-task-memory-flush
fabiocantone
Use when completing meaningful work, changing architecture, using external sources, resolving bugs, or updating project direction
SkillCFreshness: CClaude Code +9 - 0.474score
recommendation-verification
fabiocantone
Use when suggesting names, tools, libraries, services, vendors, domains, repositories, APIs, or options whose availability or current status matters
SkillCFreshness: CClaude Code +9 - 0.474score
roadmap-first-planning
fabiocantone
Use when beginning, continuing, splitting, or changing any non-trivial feature, bugfix, refactor, or multi-step coding task
SkillCFreshness: CClaude Code +9 - 0.474score
benchmark-against-superpowers
fabiocantone
Use when deciding whether this methodology is better than Superpowers, evaluating plugin quality, or running validation scenarios
SkillCFreshness: CClaude Code +9 - 0.474score
llm-wiki-memory
fabiocantone
Use when a task depends on project history, prior decisions, architecture, recurring patterns, or continuity across sessions
SkillCFreshness: CClaude Code +9 - 0.474score
plan-execution-checkpoints
fabiocantone
Use when executing a written plan, resuming a multi-step implementation, coordinating checkpoints, deciding whether to continue, pause, revise the plan, or…
SkillCFreshness: CClaude Code +9 - 0.474score
context7-research
fabiocantone
Use when a task depends on external libraries, SDKs, APIs, framework behavior, setup steps, or version-specific documentation
SkillCFreshness: CClaude Code +9 - 0.474score
field-engineering-depth
fabiocantone
Use when analyzing real code behavior, performance issues, architecture tradeoffs, production bugs, or any field task where shallow file inspection could miss…
SkillCFreshness: CClaude Code +9 - 0.474score
memory-hygiene-audit
fabiocantone
Use when checking project wiki health, roadmap freshness, source ledger quality, decision drift, session-state continuity, or cross-chat handoff quality
SkillCFreshness: CClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 1
- Last commit (this folder)
- May 26, 2026 (5 months ago)
- Last release
- May 27, 2026 (5 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/fabiocantone/codersteroids
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from fabiocantone
- 0.474score
field-engineering-depth
fabiocantone
Use when analyzing real code behavior, performance issues, architecture tradeoffs, production bugs, or any field task where shallow file inspection could miss…
SkillCFreshness: CClaude Code +9 - 0.474score
memory-hygiene-audit
fabiocantone
Use when checking project wiki health, roadmap freshness, source ledger quality, decision drift, session-state continuity, or cross-chat handoff quality
SkillCFreshness: CClaude Code +9 - 0.474score
chat-continuity
fabiocantone
Use when starting a new chat, resuming work from another chat, handing off work, or preserving project context across conversations
SkillCFreshness: CClaude Code +9 - 0.474score
context7-research
fabiocantone
Use when a task depends on external libraries, SDKs, APIs, framework behavior, setup steps, or version-specific documentation
SkillCFreshness: CClaude Code +9 - 0.474score
code-review-discipline
fabiocantone
Use when requesting code review, reviewing completed work, receiving review feedback, evaluating reviewer suggestions, or deciding whether a change is ready…
SkillCFreshness: CClaude Code +9 - 0.474score
observability-logging
fabiocantone
Use when diagnosing bugs, performance issues, production behavior, flaky systems, or any task where logs, metrics, traces, profiles, or structured runtime…
SkillCFreshness: CClaude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get branch-workspace-lifecycle running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
branch-workspace-lifecycle skill. Agent Mart. https://agentmart.dev/skills/fabiocantone-branch-workspace-lifecycle (last updated 2026-10-24)