Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
This skill reconstructs prior AI coding-agent sessions from transcripts, exports, or summaries across Claude Code, Codex, Cursor, Antigravity, and OpenCode. It extracts tasks, records provenance, and validates claims against the current workspace before continuing work. It helps developers who move between platforms, team leads reviewing agent output, and automation engineers who need clean handoffs between sessions.
WHO IT'S FOR
Cross-platform AI agent users
continuing or auditing coding sessions across multiple agents
Engineering team leads / reviewers
reviewing prior agent session work
AI automation engineers
rebuilding context from handoff transcripts across sessions
AI session auditors
verifying agent actions, provenance, and evidence integrity
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Installs: Install the agent-session-resume skill into the current project for Claude Code, Codex, Cursor, and OpenCode without prompts. · claude-code, codex, cursor, opencode
Before you start
Run from: Current project
In a terminal, run the documented non-interactive command for the four named agents.
npx skills add hacktivist123/agent-session-resume -a claude-code -a codex -a cursor -a opencode -y
After installation, restart each agent or start a new session so it loads the current instructions.
A fixture source containing a prior coding-agent transcript, handoff, chat export, session summary, or saved artifact set; replace “this fixture” by attaching it or identifying its path.
Use agent-session-resume on this fixture. Read the full source, reconstruct context, classify tasks as DONE, PARTIALLY DONE, or NOT DONE, then state the next action. Do not edit files.