aictx
Repo-local continuity for AI coding agents — preserve execution state across sessions
Free - Free and open-source (pip install aictx). Source available on GitHub.
About aictx
AICTX preserves the last useful execution state for AI coding agents so they can resume work across sessions and tools. It stores current goals, hypothesis, relevant files, next actions, decisions, failures, validation evidence and repo context, plus stale-context quality signals. Delivered as a pip-installable CLI and local control plane (local MCP) with generated agent instructions, it supports Codex, Claude, GitHub Copilot and other coding agents to resume → work → finalize without re-onboarding or repeating failures.
Key Features
- Repo-local execution state: goals, next actions, relevant files and verification evidence
- Tracks decisions, handoffs and failed commands to avoid repeat mistakes
- CLI and pip package for local installation and integration into agent workflows
- Local MCP and generated agent instructions to instruct agents to resume→work→finalize
- Quality signals and validation evidence to surface stale or unverified continuity
Who is this for?
Software engineers and teams using AI coding assistants (Codex, Claude, Copilot) who need cross-session continuity and reproducible agent workflows.
Comments (0)
No comments yet
Be the first to share your thoughts!
Sign in to join the conversation
Sign InStats
Categories
Pricing
FreeFree and open-source (pip install aictx). Source available on GitHub.
Platforms
Tech Stack
Tags
Related Products
Sora AI
Create stunning videos from text in minutes
Leonardo AI
All-in-one AI art platform with superior control
OriginBrief
Automated, source-backed weekly research and regulatory monitoring.
Nexly
Privacy-first analytics with AI-powered insights
Shoturo
AI photo editor for car listings with background and lighting fixes