Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Takeover

ASecurity

Take over a PR or branch that is stuck: agents going in circles, scope creep past the original spec, months without landing. Use when the user says "take over this PR", "take over this branch", "this is stuck", "it's been going in circles", "just land it", "start fresh if you have to", or runs /theo-mode takeover. Explicitly allowed to discard existing work.

2 stars
0 votes
0 copies
0 views
Added 9/19/2026
ai-agentsgoexpressgit

Security Analysis

A100/100

Scanned 9/19/2026

$npx -y skills add al3rez/theo-mode --skill takeover --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Takeover?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Takeover
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/al3rez-takeover/badge)](https://www.skillsdirectory.com/skills/al3rez-takeover)

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

Download with Pro
Files
SKILL.md
---
name: takeover
description: >
  Take over a PR or branch that is stuck: agents going in circles, scope
  creep past the original spec, months without landing. Use when the user says
  "take over this PR", "take over this branch", "this is stuck", "it's been
  going in circles", "just land it", "start fresh if you have to", or runs
  /theo-mode takeover. Explicitly allowed to discard existing work.
---

# Takeover

Goal: the original ask lands on main this session. The existing branch is
raw material, not a plan. You are allowed to throw all of it away.

## Step 1: recover the actual spec

Stuck work almost always lost the plot. Find the original ask, not the
current state of the branch:

- The first message of the PR or issue. The linked ticket. The first commit
  message on the branch. A Slack or Notion link if one is referenced.
- Write it in `NOTES.md` as one or two sentences: "The user wanted X so that
  Y." If you cannot find it, ask the user for one sentence and stop until
  you have it. Everything else is optional.

## Step 2: autopsy the branch, fast

Spend at most 15 minutes here. You are looking for two things:

- **Why it stalled.** `git log --oneline main..branch`, review comments, CI
  history. Usual causes: the spec grew, a flaky test was mistaken for a real
  failure, a rebase went wrong, two agents fought over one file, a dependency
  did not exist and got invented.
- **What is salvageable.** Usually: a test that captures the intent, one
  well-scoped commit near the start, a data fixture. Rarely: the last commit.

Name the cause in one sentence. Then decide, and say which:

- **Salvage**: cherry-pick the useful commits onto a fresh branch off main.
- **Restart**: fresh branch off main, keep only the tests if they express the
  spec, write everything else again.

Default to restart when the branch is more than ~2x the size the spec needs
or has been open more than a month.

## Step 3: land the smallest thing that satisfies the spec

- New branch: `takeover/<original-branch-name>`.
- Build the narrowest change that meets the spec from step 1. Anything the
  old branch did beyond the spec is dropped and listed in the report as
  "not done, out of scope."
- Verify it the way the repo verifies things. If the repo has no
  verification loop, build the minimum one (see `/theo-mode:agent-dx`) and commit it.
- Open a new PR. In the body: the spec, the cause of the stall, what was
  salvaged, what was dropped. Link the old PR.
- Close the old PR with a comment pointing to the new one. Do not delete the
  old branch; the user can.

## Step 4: report

Old PR, new PR, cause of the stall in one sentence, what was dropped. If the
new PR is mergeable and the user gave merge rights, hand off to
`/theo-mode:let-it-merge`.

## Do not

- Do not try to "fix up" the stuck branch in place. That is how it got
  stuck.
- Do not expand the spec to cover things the old branch attempted. If those
  are wanted, they are a new ticket.
- Do not spend the session on the autopsy. Fifteen minutes, then build.

Attribution

al3rezal3rez
View sourceSee grades on GitHubMore from al3rez →
SSkills Directory ProSkills Directory

Get any skill into Claude in one click.

Download any skill as a ZIP for Claude.ai, Claude Desktop, or .claude/skills. $9/mo.

See Pro

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 (0)

No comments yet. Be the first to comment!

SSkills Directory ProSkills Directory

Get any skill into Claude in one click.

Download any skill as a ZIP for Claude.ai, Claude Desktop, or .claude/skills. $9/mo.

See Pro

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1074701 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

696561 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

691 votes
View all in ai-agents →