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.
npx skills add Maximumsoft-Co-LTD/claude-flightdeck --skill recovergit clone --depth 1 https://github.com/Maximumsoft-Co-LTD/claude-flightdeckWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/maximumsoft-co-ltd/claude-flightdeck/recover)<a href="https://agentmods.dev/skills/maximumsoft-co-ltd/claude-flightdeck/recover"><img src="https://agentmods.dev/badge/skills/maximumsoft-co-ltd/claude-flightdeck/recover.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00104 | $0.02250 |
| Opus 5 | $0.00052 | $0.01125 |
| Sonnet 5 | $0.00021 | $0.00450 |
| Haiku 4.5 | $0.00010 | $0.00225 |
Grade A, and why
recover scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 243 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/recover — Safe recovery from broken git state
Walks the operator through a deliberate, non-destructive recovery
from common failure modes that follow /work (parallel fanout), hand-edit
mistakes, or interrupted merges.
Canonical reference: docs/playbooks/failure-recovery.md.
Core principle: git reflog first; destructive operations
(reset --hard, clean -fd, branch -D) only after explicit
confirmation per step.
Token budget (MANDATORY)
git status -s+git worktree list+git stash list+git log --oneline -5is enough state for 90% of recoveries. Do NOTgit diffeverything up front.git reflog | head -30only when you need to find a lost SHA.- Read FOLLOWUPS.md before writing to it; otherwise no other file reads.
- For each recovery branch, the destructive command runs only after the user types literal confirmation ("yes" or the exact SHA).
Step 1 — Inventory current state
Run all four in one shell turn (read-only):
git -C "$PWD" status -s
git -C "$PWD" worktree list
git -C "$PWD" stash list
git -C "$PWD" log --oneline -5
Plus, if the user mentioned a specific branch / SHA / file, also:
git -C "$PWD" reflog | head -30
Report back what you see — do NOT act yet. The operator may have context (e.g. "yes that worktree is mine, abandoned") that changes what's safe.
Step 2 — Classify the failure (user confirms)
Match the inventory against one of the named classes. Present the match to the user with a one-line summary and ask them to confirm.
| Class | Signals |
|---|---|
| Partial dispatch | git status shows changes from one path but not others declared by /work; some worktrees committed and others didn't |
| Orphan worktree | git worktree list shows worktrees with no recent activity / branches the user doesn't recognize |
| Dirty merge | git status -s shows UU / AA markers, or a MERGE_HEAD file exists in .git/ |
| Accidental main commit | git log --oneline -1 on main / dev / integration shows a commit you didn't intend there |
| Lost branch | A branch the user expected to find is gone (git branch --list <name> empty); SHA is recoverable via reflog for ~90 days |
| Stuck rebase | .git/rebase-apply/ or .git/rebase-merge/ exists |
| Detached HEAD with work | git status shows "HEAD detached at " + uncommitted work |
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.
- 8d ago First seen · 243 lines · 104 tokens per session scan A 3fb04f5434b3
recover is a skill published in the GitHub repository Maximumsoft-Co-LTD/claude-flightdeck (2 stars, last pushed 3mo ago), licensed MIT. It adds 104 tokens to every session and 2,250 once invoked, about $0.0005 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.
Other skills, from other repositories
pr-triage
4-phase PR backlog management with audit, deep code review, validated comments, and optional worktree setup. Use when triaging pull requests, catching up on pending code reviews, or managing a backlog of open PRs. Args: 'all' to review all, PR numbers to focus (e.g. '42 57'), 'en'/'fr' for language, no arg = audit…
campaign-commit
Choose the correct commit command in a camp, also called a campaign. Use when you are about to commit and need to select camp commit, camp p commit, fest commit, or intentional root pointer sync via camp refs-sync.
fest-execution
Execute active festival tasks. Use when finding the next task, marking tasks completed/blocked/reset, committing with festival traceability, advancing workflow steps, and validating sequence progress.
git-ai-archaeology
Analyze AI config evolution in a git repo. Use when mapping AI adoption history, finding when configs were first introduced, charting commit velocity by month, or identifying maturity phases in a project's AI tooling.
land-and-deploy
Merge PR, wait for CI, verify deploy, run canary. The complete landing pipeline.
camp-projects
Manage a camp's projects. Use when committing inside projects/, deciding status/pull/push scope (root vs submodule vs all), or creating/removing project worktrees.