pflow-backlog

If backlog-add.sh or backlog-remove.sh fails (non-zero exit or "status":"error" in its JSON) print ⚠️ <error.message> and stop — except a slug error from backlog-add.sh: invent another slug and retry (max 3 retries).

Steps

1. Clarify. You MUST know what to record and why. If the request is ambiguous, ask — never assume.

2. Write the task. Plain text for people, in the project's documentation language (from AGENTS.md/CLAUDE.md or the existing docs; else the conversation language). No title, frontmatter, status, lists or emphasis. State what must be done and only the technical detail that helps. At most 10 non-blank lines, each at most 140 characters — add line breaks instead of long lines.

3. Slug. Invent a latin kebab-case slug from the task, at most 100 characters.

4. Add. Run:

   .agents/skills/pflow-backlog/scripts/backlog-add.sh --slug "SLUG" <<'__PFLOW_BACKLOG_EOF__'
   <task text>
   __PFLOW_BACKLOG_EOF__