
Recover
Resume an interrupted agent coding session with lifecycle phase, issue tracker state, and git context in one dashboard.
Overview
Recover is an agent skill most often used in Build (also Ship, Operate) that reconstructs where your session left off from RPI phase state, bd issues, and git.
Install
npx skills add https://github.com/boshu2/agentops --skill recoverWhat is this skill?
- Reads `.agents/rpi/phased-state.json` to report the current RPI lifecycle phase
- Surfaces in-progress and ready issues from bd for claimed-work continuity
- Shows branch, recent commits, and uncommitted changes from git
- Supports `--json` for machine-readable recovery state
- Hexagon driving-adapter: consumes bd + RPI; can emit `.agents/rpi/*.md` artifacts
- 4 scenario areas in executable spec (phase, bd, git, --json)
Adoption & trust: 795 installs on skills.sh; 384 GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
You reopened the repo after a break and cannot tell which RPI phase, bd claims, or git changes were in flight.
Who is it for?
Solo builders using AgentOps RPI + bd who pause mid-session and need a fast state-of-the-world before the agent continues coding.
Skip if: Greenfield repos with no `.agents/rpi/phased-state.json`, no bd workflow, or teams that only need generic `git status` without agent lifecycle mapping.
When should I use this skill?
An agent resumes after a break, compaction, or handoff and must reconstruct in-progress session context from RPI, bd, and git.
What do I get? / Deliverables
You get a recovery dashboard or JSON snapshot so you continue claimed work instead of re-deriving context from scratch.
- Recovery dashboard
- Optional JSON recovery payload
- Optional `.agents/rpi/*.md` updates per adapter
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
AgentOps recovery is filed under Build because it reattaches the agent to an in-flight codebase and RPI workflow—the moment you sit back down to implement or fix, not when you are still researching the idea. agent-tooling is the canonical shelf for skills that manage session state, phased specs, and bd-claimed work rather than shipping features or marketing.
Where it fits
After lunch, run recover to see RPI phase and which bd issue was claimed before continuing the feature.
Resume a PR-fix session by surfacing uncommitted changes and ready bd items tied to the review.
Pick up hotfix work by confirming branch and recent commits alongside in-progress tracker state.
How it compares
Use instead of manually reading phased-state, bd, and git in separate commands when you already standardize on AgentOps.
Common Questions / FAQ
Who is recover for?
Indie and solo developers who run long Claude Code or similar agent sessions with boshu2/agentops RPI phased state and bd for issue claims.
When should I use recover?
At session start after compaction or time away during Build implementation, Ship fix cycles, or Operate iteration when you need phase + bd + git in one view.
Is recover safe to install?
It reads local project state (git, RPI, bd); review the Security Audits panel on this Prism page before trusting it in sensitive repos.
SKILL.md
READMESKILL.md - Recover
# Executable spec for the /recover skill — session-context recovery (driving-adapter). # /recover reconstructs where a session left off: the rpi lifecycle phase from # .agents/rpi/phased-state.json, the claimed/ready work from bd, and recent git state — rendered # as a recovery dashboard (or JSON). Hexagon: driving-adapter; consumes bd + rpi; produces # .agents/rpi/*.md. (soc-qk4b) Feature: Recover reconstructs in-progress session context As an agent resuming after a break or compaction I want the prior session's lifecycle phase and claimed work surfaced So that I continue where I left off instead of starting cold Scenario: the rpi lifecycle phase is detected When /recover runs Then it reads .agents/rpi/phased-state.json (when present) and reports the current phase Scenario: claimed and ready work is surfaced from bd When /recover runs Then it reports in-progress and ready issues from bd Scenario: recent git state is included When /recover runs Then it shows the current branch, recent commits, and uncommitted changes Scenario: machine-readable output on demand When /recover --json runs Then it emits the recovery state as structured JSON #!/usr/bin/env bash set -euo pipefail SKILL_DIR="$(cd "$(dirname "$0")/.." && pwd)" PASS=0; FAIL=0 check() { if bash -c "$2"; then echo "PASS: $1"; PASS=$((PASS + 1)); else echo "FAIL: $1"; FAIL=$((FAIL + 1)); fi; } check "SKILL.md exists" "[ -f '$SKILL_DIR/SKILL.md' ]" check "SKILL.md has YAML frontmatter" "head -1 '$SKILL_DIR/SKILL.md' | grep -q '^---$'" check "SKILL.md has name: recover" "grep -q '^name: recover' '$SKILL_DIR/SKILL.md'" check "SKILL.md mentions compaction" "grep -qi 'compaction' '$SKILL_DIR/SKILL.md'" check "SKILL.md mentions context recovery" "grep -qi 'context.*recovery\|recovery.*context\|recover.*context' '$SKILL_DIR/SKILL.md'" check "SKILL.md has tier: session when metadata is present" "! grep -q '^metadata:' '$SKILL_DIR/SKILL.md' || grep -q '^[[:space:]]*tier:[[:space:]]*session' '$SKILL_DIR/SKILL.md'" echo ""; echo "Results: $PASS passed, $FAIL failed" [ $FAIL -eq 0 ] && exit 0 || exit 1 --- name: recover description: Recover session context. practices: - sre - legacy-code-seams - pragmatic-programmer hexagonal_role: driving-adapter consumes: - bd - rpi produces: - .agents/rpi/*.md context_rel: [] skill_api_version: 1 context: window: inherit intent: mode: none intel_scope: none metadata: tier: session dependencies: [] output_contract: 'stdout: recovered context summary' --- # /recover — Context Recovery After Compaction > **Purpose:** Help you get back up to speed after context compaction. Detects in-progress work (RPI runs, evolve cycles), loads relevant knowledge, and summarizes what you were doing and what's next. AgentOps 3.0's default recovery path is explicit (`ao session bootstrap`, `ao codex start` / `ao codex stop`); Codex native hooks are opt-in compatibility, not assumed. **YOU MUST EXECUTE THIS WORKFLOW. Do not just describe it.** **CLI dependencies:** gt, ao, bd — all optional. Shows what's available, skips what isn't. --- ## Quick Start ```bash /recover # Full recovery dashboard /recover --json # Machine-readable JSON output ao codex status # Codex hookless lifecycle health ao codex start # Rebuild startup context explicitly in Codex ``` --- ## Execution Steps ### Step 1: Detect In-Progress Sessions (Parallel) Run ALL of the following in parallel bash calls: **Call 1 — RPI Phased State:** ```bash if [ -f .agents/rpi/phased-state.json ]; then echo "=== RPI_STATE ===" cat .agents/rpi/phased-state.json else echo "RPI_STATE=NONE" fi ``` **Call 2 — Evolve Cycle History:** ```bash if [ -f .agents/evolve/cycle-history.jsonl ]; then echo "=== EVOLVE_STATE ===" tail -3 .agents/evolve/cycle-history.jsonl else echo "EVOLVE_STATE=NONE" fi ``` **Call 3 — Git Recent Changes:** ```bash echo "=== GIT_STATUS ===" git status --short echo "=== G