Pi Durable
- ID
- 31054
- Status
- summarized
- Published
- 02 Oct 2026, 3:24 AM
- Fetched
- 02 Oct 2026, 3:20 PM
- Provider
- Hacker News
- Category
- dev-community
- Original URL
- https://earendil.com/posts/pi-durable/
- Source URL
- https://hnrss.org/best
Summary
- Score
- 6.0
- Created
- 02 Oct 2026, 3:21 PM
- Tags
- Audience
- developersai_agent_usersai_ml_learners
What happened
Earendil Engineering shipped Pi 1.0 alongside an experimental new package, Pi Durable, which it describes as a harness for long-running agents that survive internal and external failures and can be steered by multiple humans at once. Pi Durable is a framework for building any agentic application (coding agents included) rather than a replacement for the Pi coding agent, and it shares pi-ai and the project's minimalism principles with it. Earendil states the entire source without tests is roughly 15,000 lines — about 150,000 tokens with GPT and 250,000 with Claude, with storage backends alone accounting for 3,000 lines agents can usually skip.
Why it matters
The concrete claim worth acting on is the code-size-to-token ratio: a ~15k-line harness is ~150k GPT tokens / ~250k Claude tokens, which Earendil frames as something an agent can read to understand and modify the system. If you build agent tooling, that is a design constraint you can copy — keep the core small enough to fit in context and let agents skip the 3,000-line storage backends. Note that Pi Durable is explicitly experimental and the post offers no benchmarks or failure-recovery evidence, so treat the durability claims as unverified until you test them yourself. There is no Malaysia or Southeast Asia angle in this text.
Discussion angle
What actually breaks when a coding agent runs for days — and is 'small enough for an agent to read' (15k lines) a real architectural constraint worth adopting, or just a nice number for a launch post?