Researchers propose Phronix, a capability-centric learning paradigm for autonomous agents
Phronix enables agentic harnesses to acquire, validate, and refine reusable capabilities from their own executions, reducing solution non-determinism and improving reliability.
1 source · cross-referenced
- Phronix introduces a predictor-validator-corrector loop for agents to autonomously learn and refine reusable capabilities from tool executions.
Scientific computing increasingly relies on large language model-based agents for tasks such as instrumentation, analysis, and workflow porting, but their effectiveness depends heavily on manually written skills and curated memory, which limits autonomous learning.
Researchers introduce Phronix, a capability-centric learning paradigm that enables agentic harnesses to acquire, validate, and refine reusable capabilities from their own executions through a predictor-validator-corrector loop.
Phronix distills every observed interaction into a capability store and reattempts failing tasks rather than abandoning them, solving recurring tasks on the first attempt in 93–95% of cases across every model evaluated.
The paradigm turns an agent’s own execution into 7.6–30.8 durable capabilities per task without user instruction, and the number of new capabilities created falls below two once the needed capability already exists.
Warm-start runs with Phronix consume 0.38–0.40 million tokens per task, which is three to four times less than a single baseline attempt, while cold-start costs remain within the baseline token budget.
On the 1000 Genome Pegasus workflow, Phronix passes all 11 correctness checks compared to six for the baseline harness, and it increases the task-solving rates of two cheaper models from 29% and 27% to 91% and 87% on SWE-bench tasks, matching the performance of the most expensive evaluated model.
- Nov 11, 2026 · Latent Space — swyx
OpenAI demonstrates voice steering, cloud handoffs, and personal agents for Codex at AI Engineer Code Summit
Trust79 - Nov 11, 2026 · The Verge — AI
OpenAI to launch verified AI Business Agents in ChatGPT for US e-commerce and service businesses
Trust79 - Nov 10, 2026 · Latent Space — swyx
NaiveAI releases Naive-N0.5-Flash coding model with ultrafast token throughput
Trust74