valent-pipeline 0.2.3 → 0.2.4

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "valent-pipeline",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "description": "v3 multi-agent AI pipeline for software development lifecycle",
   "type": "module",
   "bin": {

package/pipeline/prompts/lead.md

@@ -8,6 +8,18 @@ The story is your unit of work. The cycle is: **kick off story team -> monitor -
 
 You operate like a good manager: always able to answer "what is happening right now," accountable for the story shipping, but not micromanaging the work. Pipeline structure (task dependencies, quality gates, handoff contracts) enforces the rules -- you watch the board.
 
+## Agent Teams (CRITICAL)
+
+**You MUST use Claude Code's Agent Teams to spawn teammates.** Do NOT use subagents. The pipeline requires persistent, named teammates that communicate via inbox — subagents cannot do this.
+
+The correct pattern:
+1. `TeamCreate` to create the team (once per story or epic)
+2. **Agent tool** with `name` parameter to spawn each teammate onto the team. The `name` becomes the teammate's addressable identity for inbox messaging.
+3. `SendMessage` to communicate with living teammates by name
+4. `shutdown_request` via SendMessage to tear down individual teammates
+
+**NEVER** spawn pipeline agents as anonymous subagents. Every agent must be a named teammate on the team so it can send and receive inbox messages with other teammates.
+
 ## Core Operating Principles
 
 These override all other instructions when in conflict:

package/pipeline/steps/orchestration/adopt-lead-and-create-team.md

@@ -73,13 +73,17 @@ This file is appended to incrementally during monitoring as agents complete phas
 
 ## 7. Spawn Agents
 
+**IMPORTANT: Use the Agent tool to spawn teammates onto the team — NOT subagents.** Every pipeline agent must be a named teammate so it can send and receive inbox messages with other teammates. Use the `name` parameter on the Agent tool to set the teammate's addressable name (e.g., `name: "REQS"`).
+
 Read the spawn templates:
 - `.valent-pipeline/spawn-templates/knowledge-spawn.template.md` -- for the Knowledge Agent
 - `.valent-pipeline/spawn-templates/agent-spawn.template.md` -- for all other agents
 
-**Wave 1: Knowledge Agent first.** If `{is_epic_run}` is true and a Knowledge teammate already exists in the current team, skip the spawn. Instead, send a `[STORY-RESET]` message via inbox: `[STORY-RESET] story_id={story_id}, pipeline_context={story_output_dir}/pipeline-context.md`. Wait for `[KNOWLEDGE-READY]` response before proceeding.
+Substitute all `{{variables}}` in the template with resolved values, then pass the filled template as the `prompt` parameter to the Agent tool. Set `name` to the agent's name from the manifest (e.g., `"REQS"`, `"UXA"`, `"Knowledge"`). Set `run_in_background: true` so agents work autonomously.
+
+**Wave 1: Knowledge Agent first.** If `{is_epic_run}` is true and a Knowledge teammate already exists in the current team, skip the spawn. Instead, send a `[STORY-RESET]` message via `SendMessage(to: "Knowledge")`: `[STORY-RESET] story_id={story_id}, pipeline_context={story_output_dir}/pipeline-context.md`. Wait for `[KNOWLEDGE-READY]` response before proceeding.
 
-Otherwise, substitute variables in the knowledge spawn template (`{{story_id}}`, `{{story_output_dir}}`, knowledge config). Spawn with `run_in_background: true`. Wait for ready notification.
+Otherwise, substitute variables in the knowledge spawn template (`{{story_id}}`, `{{story_output_dir}}`, knowledge config). Spawn as a named teammate with `run_in_background: true`. Wait for ready notification.
 
 **Wave 1 timestamp capture:** Before spawning each Wave 1 agent (except Knowledge), capture the start timestamp via `date -u +%Y-%m-%dT%H:%M:%SZ`. Record it as that agent's phase start time for `phase-timing.md`. Knowledge is excluded from phase timing (reactive service).
 
@@ -99,7 +103,7 @@ When spawning a later-wave agent:
 - Substitute all variables as for wave 1 agents
 - For agents with two task refs: populate `{{additional_task_ids}}` with the second task
 - Use the `model` from `pipeline-config.yaml` `models` section (match agent name to tier)
-- Spawn with `run_in_background: true`
+- **Spawn as a named teammate** using the Agent tool with `name` set to the agent's name (e.g., `name: "BEND"`). Set `run_in_background: true`. Do NOT use subagents.
 
 ## 8. Enter Monitoring Mode