SubMiner/AGENTS.md

<CRITICAL_INSTRUCTION>

BACKLOG WORKFLOW INSTRUCTIONS

This project uses Backlog.md MCP for all task and project management activities.

CRITICAL GUIDANCE

• If your client supports MCP resources, read backlog://workflow/overview to understand when and how to use Backlog for this project.

• If your client only supports tools or the above request fails, call backlog.get_workflow_overview() tool to load the tool-oriented overview (it lists the matching guide tools).

• First time working here? Read the overview resource IMMEDIATELY to learn the workflow

• Already familiar? You should have the overview cached ("## Backlog.md Overview (MCP)")

• When to read it: BEFORE creating tasks, or when you're unsure whether to track work

These guides cover:

• Decision framework for when to create tasks
• Search-first workflow to avoid duplicates
• Links to detailed guides for task creation, execution, and finalization
• MCP tools reference

You MUST read the overview resource to understand the complete workflow. The information is NOT summarized here.

</CRITICAL_INSTRUCTION>

Subagent Coordination Protocol (docs/subagents/)

Purpose: multi-agent coordination across runs; single-agent continuity during long runs.

Layout:

• docs/subagents/INDEX.md (active agents table)
• docs/subagents/collaboration.md (shared notes)
• docs/subagents/agents/<agent_id>.md (one file per agent)
• docs/subagents/archive/<yyyy-mm>/ (archived histories)

Required behavior (all agents):

1. At run start, read in order:
   • docs/subagents/INDEX.md
   • docs/subagents/collaboration.md
   • your own file: docs/subagents/agents/<agent_id>.md
2. Identify self by stable agent_id (runner/env-provided). If missing, create own file from template.
3. Maintain alias (short human-readable label) + mission (one-line focus).
4. Before coding:
   • record intent, planned files, assumptions in your own file.
5. During run:
   • update on phase changes (plan -> edit -> test -> handoff),
   • heartbeat at least every HEARTBEAT_MINUTES (default 5),
   • update your own row in INDEX.md (status, last_update_utc),
   • append cross-agent notes in collaboration.md when needed.
6. Write limits:
   • MAY edit own file.
   • MAY append to collaboration.md.
   • MAY edit only own row in INDEX.md.
   • MUST NOT edit other agent files.
7. At run end:
   • record files touched, key decisions, assumptions, blockers, next step for handoff.
8. Conflict handling:
   • if another agent touched your target files, add conflict note in collaboration.md before continuing.
9. Brevity:
   • terse bullets; factual; no long prose.

Suggested env vars:

• AGENT_ID (required)
• AGENT_ALIAS (required)
• HEARTBEAT_MINUTES (optional, default 20)