app-recover

A command for recovering an interrupted development run by checking its ledger, board, worktrees, and active wave before restarting. A worktree is a separate working copy, and a wave is a group of coordinated changes.

In plain words
What is it for?
Use it after a failed or interrupted run to find live attempts and leases, reconcile recorded work with the board and worktrees, and restart only from a known state.
Why use it?
A stopped run may leave tickets claimed, files leased, changes partly merged, or a manager lease active. Recovery checks each persisted state so a restart does not duplicate or discard work.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/vmobifystudio/app-dev-team/app-recover
Clone the repo
git clone --depth 1 https://github.com/vmobifystudio/app-dev-team
Per session 25 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,421 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00025 $0.01421
Opus 5 $0.00013 $0.00711
Sonnet 5 $0.00005 $0.00284
Haiku 4.5 $0.00003 $0.00142

Measured 2d ago against content hash f904c644183d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

app-recover scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/app-recover.md · 126 lines

How it starts

The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/app-recover — recover an interrupted run

Run / attempt: $ARGUMENTS

This is the highest-stress command in the studio and it used to be eleven lines of prose invoking nothing. A run dies mid-flight with worktrees leased, tickets claimed, a wave half-merged and a lease nobody released — and the instruction was to go and look. Every one of those states already has a script that can answer it definitively; none of them was called. Meanwhile /app-run-status calls run-doctor, orchestrator round calls four checks, and the command for the worst moment called zero.

Work the four questions in order. Each is a command, and each has a three-way answer. Do not skip to step 5 because an earlier step "looked fine" — a run that died is precisely the state in which looking fine is cheap.

1. What was running, and is anything still holding a lease?

node "${CLAUDE_PLUGIN_ROOT}/scripts/run-doctor.mjs" --root .
node "${CLAUDE_PLUGIN_ROOT}/scripts/run-ledger.mjs" list --root .
  • Exit 0 → no attempt is live. Continue.

  • A live lease → do not start a replacement. Either the attempt is still running somewhere, or it died holding the lease. Establish which before writing anything:

    node "${CLAUDE_PLUGIN_ROOT}/scripts/run-ledger.mjs" abandon --run <id> --attempt <n> \
      --detail "<what actually happened — an interruption is a fact, not an inference>"
    

    abandon is a terminal record and the ledger is append-only. Never edit or re-anchor it by hand: a broken chain is evidence, and repairing it destroys the only account of what happened.

2. Does the board still describe reality?

node "${CLAUDE_PLUGIN_ROOT}/scripts/board-doctor.mjs" docs/31-board.md
node "${CLAUDE_PLUGIN_ROOT}/scripts/merge-reconcile.mjs" --root .

merge-reconcile is the one that matters here, and read its wording carefully — it now distinguishes two states that look identical from the outside:

  • AWAITING INTEGRATION → merge-gated, carrying verified_static, wave not yet run. Nothing is wrong. Go to step 4 and land the wave.
  • BLOCKED — claims integrated code that is NOT in <base> → a ticket says its code shipped and git disagrees. That is a real inconsistency and the likeliest thing a crash leaves behind. Correct the record by appending, never by editing:

Read the full file on GitHub · 126 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 126 lines · 25 tokens per session scan A f904c644183d

Subscribe to this mod's changes

app-recover is a command published in the GitHub repository vmobifystudio/app-dev-team (4 stars, last pushed 23d ago), licensed MIT. It adds 25 tokens to every session and 1,421 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.