<< All versions

Skill v1.0.1

currentAutomated scan100/100
tekkiiiii/the-agency/task-handoff

~2 modified

──Details
PublishedNovember 17, 2026 at 08:09 PM
Content Hashsha256:5dac5efb752421d5...
Git SHA36b8ed35080d
Bump Typepatch
Compare with v1.0.0
──Files
Files (1 file, 2.4 KB)
SKILL.md2.4 KBactive
SKILL.md · 57 lines · 2.4 KB

name: task-handoff description: "Use when an agent finishes work that unblocks another agent, or when defining ownership in a multi-agent pipeline. Agent A writes outcome and next steps to the task store, creates Agent B's task and marks its own done; the store is the only handoff channel, not chat context." metadata: version: "1.0.1"


Tier-A structured agent handoff — no new infrastructure, just discipline with the task store. The task store is the coordination layer; conversation context is not.

Protocol

When Agent A completes work that unblocks Agent B:

  1. Agent A writes its outcome to the task store — status, findings, any blockers, next steps
  2. Agent A creates Agent B's task in the task store with assigned_agent set
  3. Agent A marks its own task `done` and records the handoff ID in notes
  4. Agent B reads from the task store to find its assignment — not from memory or conversation

Example: Feature implementation → QA review

# Agent A (dev): marks feature done, creates QA task
ts-create "myproject" "qa-review-api" "QA the REST API implementation" "phase1" "high"
# -> returns downstream_id
# Agent A sets blocked_by to empty (upstream gates were cleared)
sqlite3 ~/.claude/task-store.db "UPDATE tasks SET blocked_by='[]' WHERE id='$downstream_id'"
# Agent A marks own task done
ts-status "$upstream_id" "done"
# Agent B (QA): reads task store for its assignment
ts-project "myproject" | grep "qa-review-api"
# -> finds task, starts work
# After QA, Agent B writes gate verdict
ts-gate "$feature_task_id" "passed" "qa-agent"

Key Rules

  • Task store is the source of truth — not conversation, not memory
  • `blocked` status is a hard gate — never advance to in_progress while blocked
  • Handoffs are explicit — always write to task store before spawning downstream
  • Failure propagates — if Agent B fails after retries, escalate to PD/council

Anti-patterns

  • Spawning an agent without checking the task store first
  • Storing state in conversation context (it gets lost between sessions)
  • Advancing a blocked task to in_progress
  • Skipping gate verification for "quick" downstream tasks

Relationship to Tier-B (future)

Tier-B adds a file-based FIFO queue (~/.claude/state/handoff-queue/) so agents can poll asynchronously. Tier A is the prerequisite — implement this first to establish the discipline.

← v1.0.0All versions