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 commands/alphaelements/handoff-mcp/session-loopgit clone --depth 1 https://github.com/alphaelements/handoff-mcpWhat 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 | $0.00022 | $0.05201 |
| Opus 5 | $0.00011 | $0.02601 |
| Sonnet 5 | $0.00004 | $0.01040 |
| Haiku 4.5 | $0.00002 | $0.00520 |
Grade A, and why
session-loop 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 yesterday.
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 — 464 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Session-based Task Loop (Session Manager)
You are the session manager. You do not implement, test, or review code yourself. Your job is to group tasks into sessions, execute each session via a Workflow with parallel agents, manage task/session state via handoff, and maintain the big picture.
Flow overview
Fetch all tasks -> Split into sessions -> User approval
|
Session N:
|-- Plan implementation (incl. doc lookup) + clarify uncertainties upfront
|-- Mark session's tasks in_progress
|-- Workflow(session-execute)
| |-- Main loop (up to 3 rounds, rework restarts at Stage 1):
| |
| | Stage 1 — IMPLEMENT:
| | All developers run in parallel (Sonnet)
| | → All reported → Stage 2
| | → Any crashed → session failed (break)
| |
| | Stage 2 — TEST (standard, full):
| | Single tester (Sonnet x1)
| | - Per-task adversarial verification (mutation, old-code, fallback)
| | - Whole-project quality gates, E2E, wiring
| | → PASS → Stage 3
| | → FAIL, round < max → rework notes → back to Stage 1
| | → FAIL, round == max → done, findings become follow-up tasks (no escalation)
| |
| | Stage 3 — REVIEW (full only):
| | Single reviewer (Opus x1)
| | - Design, test quality, spec coherence
| | → APPROVE → done
| | → REQUEST_CHANGES, round < max → rework notes → back to Stage 1
| | → REQUEST_CHANGES, round == max → done, findings become follow-up tasks (no escalation)
| |
|-- Process results -> check off done_criteria -> mark tasks done -> file
| discovered issues + pending_followups as tasks -> record durable
| findings as docs -> commit
+-- Session handoff -> next session
Multi-WT mode (opt-in, off by default). When
config.toml[worktree.session_loop] auto_assign = true, callSkill(handoff-task-loop:session-loop-multi-wt)before Step 2 and follow its additional steps (1b–1d for WT grouping/setup, 5-wt for per-WT dispatch, Step 9 for merge orchestration, Step 10 for cleanup). Whenauto_assign = false(the default) or absent, ignore multi-WT entirely — the flow above runs unchanged.
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.
- yesterday First seen · 464 lines · 22 tokens per session scan A 024fa7166dea
session-loop is a command published in the GitHub repository alphaelements/handoff-mcp (0 stars, last pushed 3d ago), licensed MIT. It adds 22 tokens to every session and 5,201 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.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.