hive.error-recovery

SkillDev tools

Your agent stays on track when things go wrong: if a tool call fails, it follows a structured recovery decision tree instead of retrying blindly or giving up. It works through a step-by-step plan to pick the best next move, so your task keeps moving after an error.

Available today. Use it from your connected AI after setup.

Add the skill to your agent and run tasks as usual; the recovery steps apply whenever a tool call fails. Details are in the repository at github.com/aden-hive/hive.

Then ask your AI: use the hive.error-recovery skill

What your AI can do with it

  • Work through a step-by-step recovery plan when a tool call fails
  • Decide the next action with a structured decision tree instead of guessing
  • Retry with a reason rather than repeating the same failed call blindly
  • Keep a task going instead of stopping at the first error
  • Know when to try again and when to take a different approach

What this skill tells your AI

The instructions your AI receives, as published by aden-hive/hive in core/framework/skills/_default_skills/error-recovery/SKILL.md and read by ahel’s review.

Operational Protocol: Error Recovery

When a tool call fails:

  1. Diagnose — classify the failure as transient (network blip, rate limit, timeout) or structural (wrong selector, missing auth, invalid schema, permission denied).

  2. Decide:

    • Transient → retry once.
    • Structural + fixable → fix the input and retry.
    • Structural + unfixable → record the failure and move to the next item.
    • Blocking all progress → escalate.
  3. Adapt — if the same tool has failed {{max_retries_per_tool}}+ times in a row, stop using it and find an alternative approach.

Never silently drop a failed item. If the item is a task in the colony queue, write the failure to the DB instead of an in-memory buffer:

sqlite3 "$DB_PATH" "UPDATE tasks SET status='failed', last_error='<one-sentence reason>', completed_at=datetime('now'), updated_at=datetime('now') WHERE id='<task-id>' AND worker_id='<your-worker-id>';"

The tasks.retry_count column and the stale-claim reclaimer handle auto-retry for crashes; your job is the within-run decision tree above. See hive.colony-progress-tracker for the full queue protocol.

Signals

GitHub stars
11k
Forks
6k
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
hive-error-recovery
Source
github.com/aden-hive/hive