forge-resumelisted

This skill should be used at the start of a work session on a project that uses the Context Forge methodology — phrases like "forge-resume", "resume the project", "where did we leave off", "pick up where we stopped", "restore context", or "read the context files and continue". It restores project context tier by tier (digest and tracker first, full files only as the task requires) so work continues without drift and without burning tokens.
yerros/context-forge · ★ 0 · AI & Automation · score 75

Install: claude install-skill yerros/context-forge

# forge-resume Restore project context in one step and continue the build without re-explaining the project. This solves the "AI has no memory between sessions" problem — without paying for context the session doesn't need. Loading follows the tier system defined in `${CLAUDE_PLUGIN_ROOT}/skills/forge-resume/references/token-economy.md`. ## Argument Text after the command sets the session's focus (e.g. `/forge-resume unit 05` or `/forge-resume the payment bug`) — resume as below, then steer step 4 to that focus and load its Tier 2 file(s). No argument → standard resume to whatever is Next Up. ## What to do 0. Confirm the project is set up. Run the deterministic detector (read-only): ```bash bash "${CLAUDE_PLUGIN_ROOT}/skills/forge-init/scripts/detect.sh" ``` If the `verdict` is `SETUP`, there's nothing to resume — suggest `forge-init`. If `REPAIR` (incomplete), mention which files are missing and suggest `forge-init`'s reconcile flow, then resume with what exists. 1. **Tier 1 — always.** Read the entry point (`CLAUDE.md` or `AGENTS.md`), then `context/context-digest.md` (if present), then `context/progress-tracker.md` — stable files first, the volatile tracker last, so the unchanged prefix stays prompt-cache-friendly across sessions. If there is **no digest** (project pre-dates it), fall back to reading the six files in full, in order: project-overview, architecture, ui-context, code-standards, ai-workflow-rules, progress-tracker —