I tried SKILL.state in my coding agents and stopped after four days

Description and use

The author tested SKILL.state in coding agents but stopped after four tasks, finding that the state file added overhead while the problem it solved had largely vanished after upgrading to Opus 5.5.

Problem

Does an explicit state file improve long‑task coding agents or add unnecessary overhead?

Outcome

The SKILL.state approach improved brief clarity but introduced schema failures, invented timestamps, and linter issues; after moving to Opus 5.5 the context problem diminished, making the state file overhead unjustified.

Workflow

• 1. Create a JSON state file per task containing objective, phase, decisions, findings, and next action.
• 2. Update the state only through a CLI that validates patches against a schema.
• 3. Launch a fresh subagent for each step, providing it the skill, current state, and latest observation, with no chat history.

Evidence

SKILL.state swaps an agent's growing chat history for a state file and a fresh step each time.

Source

The briefs were good. Each subagent got a small, self-contained frame and needed no history.

Source

Patches that failed the schema.

Source

Invented timestamps.

Source

After that, the median peak context was 162K tokens, and only 2 of 48 units of work were compacted at all.

Source