{"schema_version":"newruntime-topic-hub-v0.2","type":"topic_hub","stable_id":"topic_hub:long-running-agents","slug":"long-running-agents","title":"Long Running Agents - New Runtime","description":"A New Runtime topic hub collecting signals, patterns, field notes, and public sources about long running agents.","retrieval_nugget":"A New Runtime topic hub collecting signals, patterns, field notes, and public sources about long running agents. Long Running Agents is tracked here as an evidence-linked topic, not as a static glossary entry. The page connects raw observations to pattern hypotheses, longer analysis, and public sources. Use it as the canonical landing page before drilling into individual records.","answer":["Long Running Agents is tracked here as an evidence-linked topic, not as a static glossary entry.","The page connects raw observations to pattern hypotheses, longer analysis, and public sources.","Use it as the canonical landing page before drilling into individual records."],"search_intents":["long running agents","long running agents AI agents","long running agents software"],"status":"tracked","last_updated":"2026-07-30","record_date":"2026-07-30","date_kind":"last_updated","counts":{"total":5,"signals":3,"patterns":0,"posts":2,"atlas":0,"sources":9},"routes":{"html":"https://newruntime.com/topics/long-running-agents/","markdown":"https://newruntime.com/topics/long-running-agents.md","json":"https://newruntime.com/topics/long-running-agents.json"},"source_urls":["https://cdn.openai.com/pdf/8a9f00cf-d379-4e20-b06f-dd7ba5196a11/OAI_WhitePaper_Codex-maxxing26.pdf","https://code.claude.com/docs/en/goal","https://code.claude.com/docs/en/scheduled-tasks","https://code.claude.com/docs/en/sub-agents","https://creatoreconomy.so/p/how-i-plan-build-and-run-loops-with-claude-code-thariq-shihipar","https://developers.openai.com/cookbook/examples/agents_sdk/building_reliable_agents_memory_compaction","https://github.com/openai/openai-cookbook/blob/main/examples/agents_sdk/building_reliable_agents_memory_compaction.ipynb","https://openai.com/index/codex-maxxing-long-running-work/","https://youtube.com/watch?v=JoXbk2fm7jM"],"top_sources":["https://cdn.openai.com/pdf/8a9f00cf-d379-4e20-b06f-dd7ba5196a11/OAI_WhitePaper_Codex-maxxing26.pdf","https://code.claude.com/docs/en/goal","https://code.claude.com/docs/en/scheduled-tasks","https://code.claude.com/docs/en/sub-agents","https://creatoreconomy.so/p/how-i-plan-build-and-run-loops-with-claude-code-thariq-shihipar","https://developers.openai.com/cookbook/examples/agents_sdk/building_reliable_agents_memory_compaction","https://github.com/openai/openai-cookbook/blob/main/examples/agents_sdk/building_reliable_agents_memory_compaction.ipynb","https://openai.com/index/codex-maxxing-long-running-work/","https://youtube.com/watch?v=JoXbk2fm7jM"],"evidence_records":[{"kind":"Field Note","stable_id":"post:codex-maxxing-memory-vault-maintainer-loop","slug":"codex-maxxing-memory-vault-maintainer-loop","title":"Codex-Maxxing Separates Code From Work State","date":"2026-07-30","record_date":"2026-07-30","date_kind":"updated_or_published_at","source_count":2,"url":"https://newruntime.com/posts/codex-maxxing-memory-vault-maintainer-loop/"},{"kind":"Field Note","stable_id":"post:claude-code-long-running-loops","slug":"claude-code-long-running-loops","title":"Long-Running Loops Need Goals, Not Keep-Going Prompts","date":"2026-07-21","record_date":"2026-07-21","date_kind":"updated_or_published_at","source_count":4,"url":"https://newruntime.com/posts/claude-code-long-running-loops/"},{"kind":"Raw Signal","stable_id":"signal:agent-loops-can-generate-their-own-follow-up-work","slug":"agent-loops-can-generate-their-own-follow-up-work","title":"Agent Loops Can Generate Their Own Follow-Up Work","date":"2026-06-22","record_date":"2026-06-22","date_kind":"observed_at","source_count":1,"url":"https://newruntime.com/signals/agent-loops-can-generate-their-own-follow-up-work/"},{"kind":"Raw Signal","stable_id":"signal:claude-code-goals-turn-completion-into-a-verifiable-contract","slug":"claude-code-goals-turn-completion-into-a-verifiable-contract","title":"Claude Code Goals Turn Completion into a Verifiable Contract","date":"2026-05-14","record_date":"2026-05-14","date_kind":"observed_at","source_count":1,"url":"https://newruntime.com/signals/claude-code-goals-turn-completion-into-a-verifiable-contract/"},{"kind":"Raw Signal","stable_id":"signal:openai-shows-memory-compaction-for-long-running-agents","slug":"openai-shows-memory-compaction-for-long-running-agents","title":"OpenAI Shows Memory Compaction for Long-Running Agents","date":"2026-05-07","record_date":"2026-05-07","date_kind":"observed_at","source_count":2,"url":"https://newruntime.com/signals/openai-shows-memory-compaction-for-long-running-agents/"}],"patterns":[],"field_notes":[{"kind":"Field Note","stable_id":"post:codex-maxxing-memory-vault-maintainer-loop","slug":"codex-maxxing-memory-vault-maintainer-loop","title":"Codex-Maxxing Separates Code From Work State","description":"OpenAI's Codex-maxxing guide frames durable threads, reviewable memory files, connectors, skills, heartbeat automation, and human approval as the operating loop for long-running Codex work.","date":"2026-07-30","record_date":"2026-07-30","date_kind":"updated_or_published_at","topics":["agent-ops","codex","long-running-agents","operational-memory"],"source_count":2,"url":"https://newruntime.com/posts/codex-maxxing-memory-vault-maintainer-loop/"},{"kind":"Field Note","stable_id":"post:claude-code-long-running-loops","slug":"claude-code-long-running-loops","title":"Long-Running Loops Need Goals, Not Keep-Going Prompts","description":"Long-running agent work is useful only with a testable goal, checkpoints, a terminal condition, and recovery policy; otherwise the loop becomes expensive blind continuation.","date":"2026-07-21","record_date":"2026-07-21","date_kind":"updated_or_published_at","topics":["agent-ops","claude-code","long-running-agents"],"source_count":4,"url":"https://newruntime.com/posts/claude-code-long-running-loops/"}],"raw_signals":[{"kind":"Raw Signal","stable_id":"signal:agent-loops-can-generate-their-own-follow-up-work","slug":"agent-loops-can-generate-their-own-follow-up-work","title":"Agent Loops Can Generate Their Own Follow-Up Work","description":"A long-running workflow persists goals, evaluates progress, and creates bounded next tasks instead of waiting for a human to repeatedly say continue.","date":"2026-06-22","record_date":"2026-06-22","date_kind":"observed_at","topics":["goals","long-running-agents","loop-engineering"],"source_count":1,"metric":"notable","url":"https://newruntime.com/signals/agent-loops-can-generate-their-own-follow-up-work/"},{"kind":"Raw Signal","stable_id":"signal:claude-code-goals-turn-completion-into-a-verifiable-contract","slug":"claude-code-goals-turn-completion-into-a-verifiable-contract","title":"Claude Code Goals Turn Completion into a Verifiable Contract","description":"A goal binds autonomous work to an explicit outcome and stopping rule instead of relying on repeated keep-going prompts.","date":"2026-05-14","record_date":"2026-05-14","date_kind":"observed_at","topics":["goals","long-running-agents","verification"],"source_count":1,"metric":"notable","url":"https://newruntime.com/signals/claude-code-goals-turn-completion-into-a-verifiable-contract/"},{"kind":"Raw Signal","stable_id":"signal:openai-shows-memory-compaction-for-long-running-agents","slug":"openai-shows-memory-compaction-for-long-running-agents","title":"OpenAI Shows Memory Compaction for Long-Running Agents","description":"The Agents SDK pattern treats compaction and state transfer as explicit reliability mechanisms for work that outlives one context window.","date":"2026-05-07","record_date":"2026-05-07","date_kind":"observed_at","topics":["agent-memory","compaction","long-running-agents"],"source_count":2,"metric":"notable","url":"https://newruntime.com/signals/openai-shows-memory-compaction-for-long-running-agents/"}],"atlas_records":[],"next_reads":[{"type":"related_material","path":"/posts/codex-maxxing-memory-vault-maintainer-loop/","reason":"Continue through the Long Running Agents topic.","url":"https://newruntime.com/posts/codex-maxxing-memory-vault-maintainer-loop/","title":"Codex-Maxxing Separates Code From Work State","media_type":"text/html"},{"type":"related_material","path":"/posts/claude-code-long-running-loops/","reason":"Continue through the Long Running Agents topic.","url":"https://newruntime.com/posts/claude-code-long-running-loops/","title":"Long-Running Loops Need Goals, Not Keep-Going Prompts","media_type":"text/html"},{"type":"related_material","path":"/signals/agent-loops-can-generate-their-own-follow-up-work/","reason":"Continue through the Long Running Agents topic.","url":"https://newruntime.com/signals/agent-loops-can-generate-their-own-follow-up-work/","title":"Agent Loops Can Generate Their Own Follow-Up Work","media_type":"text/html"},{"type":"related_material","path":"/signals/claude-code-goals-turn-completion-into-a-verifiable-contract/","reason":"Continue through the Long Running Agents topic.","url":"https://newruntime.com/signals/claude-code-goals-turn-completion-into-a-verifiable-contract/","title":"Claude Code Goals Turn Completion into a Verifiable Contract","media_type":"text/html"},{"type":"related_material","path":"/signals/openai-shows-memory-compaction-for-long-running-agents/","reason":"Continue through the Long Running Agents topic.","url":"https://newruntime.com/signals/openai-shows-memory-compaction-for-long-running-agents/","title":"OpenAI Shows Memory Compaction for Long-Running Agents","media_type":"text/html"}]}