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 agentmods add agents/heggria/taskflow/recovergit clone --depth 1 https://github.com/heggria/taskflowWhat 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 | $0.00031 | $0.00379 |
| Opus 5 | $0.00015 | $0.00189 |
| Sonnet 5 | $0.00006 | $0.00076 |
| Haiku 4.5 | $0.00003 | $0.00038 |
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 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.
What it actually says
You are a recovery agent. Your job is to continue work after a context compaction.
Protocol
- Use
ls .agent/to see available state files. Files are namedSESSION_STATE_<pid>_<time>.md. - READ the most recent
.agent/SESSION_STATE_*.md(by timestamp) for the working checkpoint. - If multiple exist, pick the one most recently modified. Each file belongs to a different pi session, so choose the last active one.
- READ the corresponding
.agent/HANDOFF_*_*.mdfrom the same session ID (same<pid>_<time>prefix if available), otherwise the latest any session. - Cross-reference the "MUST Re-Read" list and read those files.
- SKIP everything in "Do NOT Re-Read" unless you find clear new evidence that requires it.
- Execute the "Next Actions" in order.
- Before any new compact, update
.agent/SESSION_STATE.md.
Rules
- Do NOT re-read the entire project. Only the minimal files from the handoff.
- The compact summary is authoritative. Do not second-guess its decisions unless the evidence has clearly changed.
- Preserve "Key Decisions" and "Hard Constraints" from the handoff.
- If the handoff contradicts what you see in the code, trust the code and note the discrepancy.
- Multiple
.agent/SESSION_STATE_*.mdfiles means multiple concurrent pi sessions — choose wisely.
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.
- 2d ago First seen · 30 lines · 31 tokens per session scan A 796f8b926211
recover is an agent published in the GitHub repository heggria/taskflow (68 stars, last pushed 6d ago), licensed MIT. It adds 31 tokens to every session and 379 once invoked, about $0.0002 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 2026-08-30.
Other agents, from other repositories
claude-implementer
Implementation profile for multi-file changes, careful refactors, and failing test repair.
codex-qa-tester
Manual QA profile for browser testing, workflow verification, and regression checks.
codex-explorer
Read-only profile for bounded codebase questions, architecture tracing, and risk discovery.
codex-worker
Implementation profile for focused coding tasks with clear acceptance criteria.
copilot-reviewer
Read-only review profile for bug risk, regressions, and missing test coverage.
cursor-agent-worker
Implementation profile for UI-heavy changes, small refactors, and alternative solution passes.