Domain Retry

Retrying Failed Domains

When a domain fails during aspens doc init, the CLI shows the exact command to retry just that domain:

▲  1 domain(s) skipped: performance-tracking
  Retry just these: aspens doc init --mode chunked --domains "performance-tracking"
  Or retry all: aspens doc init --mode chunked --timeout 600

What --domains does in chunked mode

aspens doc init --mode chunked --domains "performance-tracking"

This runs a fast, targeted regeneration:

1. Scans repo — fast, no LLM
2. Skips discovery — no parallel agents, saves ~2 min
3. Skips strategy prompt — no "improve/rewrite/skip" question
4. Skips base skill — loads existing base skill for context instead of regenerating
5. Generates only the specified domain(s)
6. Skips CLAUDE.md — already exists from the first run
7. Updates hooks — re-generates skill-rules.json to include the new skill

Total time: ~1 min instead of ~9 min for a full run.

Multiple domains

Comma-separate domain names to retry several at once:

aspens doc init --mode chunked --domains "performance-tracking,github-traffic"

When to use this

• A domain timed out during the initial doc init run
• You want to regenerate a specific skill after making code changes
• You added new files to a domain and want the skill updated

Related

• How Aspens Works — full pipeline overview
• Common Recipes — other useful command combinations