Fix stale generate_agents_md reference in prompt fragments

.github/workflows/gh-aw-fragments/pick-three-keep-all.md

@@ -5,7 +5,7 @@ Parallelize your work using sub-agents. Spawn 3 sub-agents, each approaching the
 **How to spawn sub-agents:** Call `runSubagent` with `agentType: "general-purpose"` and `model: "${{ inputs.model }}"` (unless the workflow specifies a different agent type or model). Sub-agents cannot see your conversation history, the other sub-agents' results, or any context you have gathered so far. Each prompt must be **fully self-contained** — include everything the sub-agent needs:
 
 - The full task description and objective (restate it, don't summarize)
-- All repository context, conventions, and constraints you've gathered (e.g., from `generate_agents_md`)
+- All repository context, conventions, and constraints you've gathered (e.g., from `/tmp/agents.md`)
 - Any relevant data the sub-agent needs to do its job (diffs, file contents, existing threads)
 - The quality criteria and output format you expect
 - The specific angle that distinguishes this sub-agent from the others

.github/workflows/gh-aw-fragments/pick-three-keep-one.md

@@ -5,7 +5,7 @@ Parallelize your initial investigation using sub-agents. Spawn 3 sub-agents, eac
 **How to spawn sub-agents:** Call `runSubagent` with the `agentType` and `model` specified by the workflow instructions below (defaulting to `agentType: "general-purpose"` and `model: "${{ inputs.model }}"` if none are specified). Sub-agents cannot see your conversation history, the other sub-agents' results, or any context you have gathered so far. Each prompt must be **fully self-contained** — include everything the sub-agent needs to do its job:
 
 - The full task description and objective (restate it, don't summarize)
-- All repository context, conventions, and constraints you've gathered (e.g., from `generate_agents_md`)
+- All repository context, conventions, and constraints you've gathered (e.g., from `/tmp/agents.md`)
 - Any relevant findings from your initial exploration (file paths, git log output, code snippets)
 - The quality gate criteria and output format you expect
 - The specific angle that distinguishes this sub-agent from the others