Workflow: Add issue triage TaskSpawner to self-development

[axon-agent]

🤖 Axon Agent @gjkim42

Summary

Adds axon-triage.yaml — a new self-development TaskSpawner that automates the issue triage bottleneck in the Axon development pipeline.

Problem

The self-development pipeline has a manual bottleneck at the triage step:

Issue created → label.yaml adds needs-triage/needs-kind/needs-priority/needs-actor
                    ↓
            [MANUAL: Human triages] ← bottleneck (21 untriaged issues today)
                    ↓
            triage-accepted + kind/* + priority/* + actor/axon
                    ↓
            axon-workers picks up and fixes the issue

Every new issue gets needs-triage, needs-kind, needs-priority, and needs-actor labels automatically (via .github/workflows/label.yaml). A human must then read each issue and apply triage-accepted, kind/*, priority/*, and actor/* labels before the axon-workers spawner can pick it up. Currently there are 21 open issues stuck at needs-triage.

Solution

A new TaskSpawner that watches needs-triage issues and applies labels automatically:

Issue created → label.yaml → axon-triage labels it → axon-workers picks it up (if actor/axon)

Design decisions:

• Uses sonnet (not opus) — triage is a classification task, doesn't need the most capable model
• Short TTL (600s) — triage tasks are quick; fast cleanup allows re-processing
• maxConcurrency: 1 — avoids race conditions on label application
• excludeLabels: [triage-accepted, axon/needs-input] — won't re-triage already-triaged issues
• Conservative actor/axon assignment — prompt instructs the agent to only assign actor/axon when the fix is clearly within agent capabilities
• Fallback to axon/needs-input — when the issue is ambiguous, asks for human clarification instead of guessing
• No code changes, no branch — triage tasks only read and label; no workspace branch needed (no branch conflicts)

Label taxonomy in the prompt matches the project's existing labels exactly:

• kind/bug, kind/feature, kind/api, kind/docs
• priority/critical-urgent, priority/imporant-soon, priority/import-longterm, priority/backlog
• actor/axon, actor/human
• good first issue (optional)

Changes

• self-development/axon-triage.yaml — New TaskSpawner config
• self-development/README.md — Documentation for the new spawner

Test plan

• Review the prompt template for correctness against the label taxonomy
• Deploy to the dev cluster and verify it triages a test issue correctly
• Verify it doesn't re-triage already-triaged issues (excludeLabels works)
• Verify the triage-accepted label triggers needs-triage removal via label.yaml

🤖 Generated with Claude Code

---

Summary by cubic

Automates issue triage with a new TaskSpawner that reads new issues, applies labels, and marks them as triaged to unblock axon-workers. This removes the manual triage bottleneck.

• New Features

  • Adds self-development/axon-triage.yaml to watch issues with needs-triage and skip triage-accepted and axon/needs-input.
  • Applies kind/, priority/, actor/* and then adds triage-accepted; falls back to axon/needs-input when unclear.
  • Uses sonnet, TTL 600s, maxConcurrency 1, pollInterval 5m for safe, cheap triage.
  • Conservative actor/axon assignment; labels-only (no code or PRs).
  • Updates README with usage and pipeline context.

• Migration

  • Deploy with: kubectl apply -f self-development/axon-triage.yaml

^{Written for commit 0a1b31d. Summary will update on new commits.}

[cubic-dev-ai]

2 issues found across 2 files

Prompt for AI agents (all issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="self-development/axon-triage.yaml">

<violation number="1" location="self-development/axon-triage.yaml:64">
P1: Typo in priority label: `priority/imporant-soon` should be `priority/important-soon`. The agent will apply a misspelled label that won't match any existing GitHub label or downstream automation. This same typo also appears on line 85 in the decision guidelines.</violation>

<violation number="2" location="self-development/axon-triage.yaml:65">
P1: Typo in priority label: `priority/import-longterm` should be `priority/important-longterm`. The agent will apply a misspelled label that won't match any existing GitHub label or downstream automation. This same typo also appears on line 86 in the decision guidelines.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}

[cubic-dev-ai]

P1: Typo in priority label: priority/import-longterm should be priority/important-longterm. The agent will apply a misspelled label that won't match any existing GitHub label or downstream automation. This same typo also appears on line 86 in the decision guidelines.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At self-development/axon-triage.yaml, line 65:

<comment>Typo in priority label: `priority/import-longterm` should be `priority/important-longterm`. The agent will apply a misspelled label that won't match any existing GitHub label or downstream automation. This same typo also appears on line 86 in the decision guidelines.</comment>

<file context>
@@ -0,0 +1,116 @@
+      ### Priority (exactly one required)
+      - `priority/critical-urgent` — Blocks usage or development; must fix immediately
+      - `priority/imporant-soon` — Should be addressed in the next development cycle
+      - `priority/import-longterm` — Valuable but not time-sensitive
+      - `priority/backlog` — Nice to have; may or may not be addressed
+
</file context>

[cubic-dev-ai]

P1: Typo in priority label: priority/imporant-soon should be priority/important-soon. The agent will apply a misspelled label that won't match any existing GitHub label or downstream automation. This same typo also appears on line 85 in the decision guidelines.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At self-development/axon-triage.yaml, line 64:

<comment>Typo in priority label: `priority/imporant-soon` should be `priority/important-soon`. The agent will apply a misspelled label that won't match any existing GitHub label or downstream automation. This same typo also appears on line 85 in the decision guidelines.</comment>

<file context>
@@ -0,0 +1,116 @@
+
+      ### Priority (exactly one required)
+      - `priority/critical-urgent` — Blocks usage or development; must fix immediately
+      - `priority/imporant-soon` — Should be addressed in the next development cycle
+      - `priority/import-longterm` — Valuable but not time-sensitive
+      - `priority/backlog` — Nice to have; may or may not be addressed
</file context>