- Home
- Skills
- Testing & Debugging
- runledger rollback
runledger rollback
Non-destructive working-tree snapshots and selective rollbacks for AI agent coding sessions.
Free
Works with the AI Agents you already use
runledger rollback
Example session with this skill installed
I messed up the refactor in the src directory. Show me the checkpoints and help me dry-run a rollback for just the src/auth.py file from the 'Before refactor' snapshot.
- Read your context and instructions
- Compiled the runledger rollback
I have retrieved the available checkpoints and prepared the dry-run for your specific file restoration.
RunLedger Rollback Preview
| Checkpoint ID | Timestamp | Label | Status |
|---|---|---|---|
rl_8f2d1 | 2023-10-24 14:20:05 | Before refactor | Active |
rl_a9e32 | 2023-10-24 14:45:12 | Post-auth-update | Active |
Dry-run Output for src/auth.py (Checkpoint rl_8f2d1):
--- src/auth.py (Current)
+++ src/auth.py (Checkpoint rl_8f2d1)
@@ -12,8 +12,5 @@
- def validate_session(token):
- # New buggy implementation
- return True if token else False
+ def validate_session(token, secret_key):
+ return jwt.decode(token, secret_key, algorithms=["HS256"])
Execution Command:
runledger rollback rl_8f2d1 --path src/auth.py --dry-run
Safety Note:
Applying this rollback will create a safety checkpoint (e.g., rl_undo_992) allowing you to revert this specific recovery if needed.
Next steps
- Run
runledger rollback rl_8f2d1 --path src/auth.py --yesto apply the change. - Verify the fix by running your auth unit tests.
- Create a new checkpoint once the fix is confirmed.
Connects securely to your tools. The creator never sees your data.
About this skill
AI agents often make destructive edits that git stash or manual branching can't easily isolate. This skill manages non-destructive working-tree snapshots using RunLedger, allowing you to create checkpoints before refactors and restore specific files or entire trees without touching HEAD or the staging index.
What it does
- Instant Snapshots creates checkpoints of tracked and untracked files without git stashing.
- Selective Rollback restores specific file paths or directories from any historical checkpoint.
- Dry-run Previews generates diff previews of changes before applying a rollback to the working tree.
- Auto-Undo Safety generates a temporary safety checkpoint before every rollback operation automatically.
- Session Linking associates checkpoints with specific agent session IDs for better traceability.
How it works
- Checkpointing uses
runledger checkpointwith an optional label to freeze the current working tree state. - Verification lists existing snapshots via
runledger checkpointsto find the correct recovery point. - Dry-run executes a simulated rollback to verify the diff and impacted files.
- Restoration applies the rollback to the tree, creating an automatic undo point in the process.
Frameworks & tools
This skill requires Python 3.9+ and the runledger-ai CLI package. It operates on any local Git repository but functions independently of the Git index and branch pointer.
Why this beats prompting it yourself
Manual git commands often fail to capture untracked files or require messy stash management that agents frequently break. This skill provides a dedicated CLI wrapper that ensures every agent action is reversible with a single command and zero impact on your commit history.
Use cases
- Protect the working tree before large-scale automated refactoring sessions.
- Revert a single failed file edit while keeping other successful agent changes.
- Recover from agent errors that deleted untracked configuration or asset files.
- Compare current project state against a specific agent session receipt.
Known limitations
Requires the runledger-ai Python package to be installed on the system path. It does not manage remote repository state or cloud-based backups.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Testing & Debugging.
Requires RunLedger CLI installed via pip. Operates only on local filesystems. Cannot restore files explicitly ignored by .gitignore.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 3 days ago
- Free with an Agensi account
- Software
Python
Git