Source: https://internal-agents.com/agents/openai-sevbot # OpenAI — Sevbot OpenAI's internal incident-response agent, built on Codex. When an incident is detected, the bot wakes up, collects context, determines possible mitigations without executing any, and answers developers' questions in the incident Slack channel; an engineer can tell it to apply a chosen mitigation. - Approach type: Agent - Deployment stage: Deployed - Autonomy: Drafts reviewed - Evidence strength: Secondary only - Status: Internal - First reported year: 2026 - Work: On-call - Interfaces: Slack - Invocation: Event driven - Entry reviewed: 2026-09-16 ## Where people stay involved Each scope pairs its normal attention boundary with supporting evidence. See the [supervision definitions](https://internal-agents.com/definitions#supervision) for the level mapping and limits. - **incident detected -> applied mitigation** — Work product review · Level 3 ## Overview ### Summary OpenAI's internal incident-response agent, built on Codex. When an incident is detected, the bot wakes up, collects context, determines possible mitigations without executing any, and answers developers' questions in the incident Slack channel; an engineer can tell it to apply a chosen mitigation. Fact · Reported · High confidence · `openai-sevbot--summary` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/openai-sevbot-article/content.md) · Preserved content.md, lines 188-193 (pipeline step 9: respond to outages) - Supports · [2] [Certificate Transparency records for sevbot.api.openai.com](https://crt.sh/?q=sevbot.api.openai.com) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/crtsh-sevbot-certificates/content.md) · Preserved content.md, lines 13-41 (certificate table for sevbot.api.openai.com; earliest not-before 2025-11-20 at lines 37-41, with renewals through 2026-08-01; evidences a deployed OpenAI service under the Sevbot name, not its incident-response behavior) - Supports · [3] [\[SEVBot\] Optionize initialized notification tolerance (openai-agents-python pull 2765 commits)](https://github.com/openai/openai-agents-python/pull/2765/commits) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/github-sevbot-pr/content.md) · Preserved content.md, lines 21-27 (initial commit '[SEVBot] Optionize initialized notification tolerance', 219a67a, 2026-03-24, authored by an OpenAI employee and co-authored by Codex; evidences internal Sevbot-branded engineering work) ## How it works ### Supporting component The bot determines possible mitigations but never executes any; an engineer tells it to apply a specific mitigation Fact · Reported · Medium confidence · `openai-sevbot--primitives-0` Confidence reason: A linked participant or independent source reports the claim. Evidence: - Supports · [1] [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/openai-sevbot-article/content.md) · Preserved content.md, lines 190-193 (mitigation bullet list) ## Supervision evidence ### Operating model assessment Level 3 for incident detected -> applied mitigation; human attention boundary: work-product-review. Inference · Catalog judgment · High confidence · `openai-sevbot--operating-models-0` Confidence reason: The article enumerates the boundary: Sevbot proposes mitigations and never executes any; an engineer commands the application. Qualifications: - Observation date: 2026-09 Evidence: - Supports · [1] [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/openai-sevbot-article/content.md) · Preserved content.md, lines 190-193 (determines possible mitigations but never executes any; an engineer can tell it to apply a specific mitigation) ## Implementation details ### Sandbox unknown Inference · Catalog judgment · Medium confidence · `openai-sevbot--architecture-sandbox` Confidence reason: The article does not document an execution sandbox; unknown does not mean absent. Evidence: - Supports · [1] [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/openai-sevbot-article/content.md) · Preserved content.md, lines 188-195 describe Sevbot without documenting an execution sandbox ### Harness Built on top of Codex Fact · Reported · Medium confidence · `openai-sevbot--architecture-harness` Confidence reason: A linked participant or independent source reports the claim. Evidence: - Supports · [1] [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/openai-sevbot-article/content.md) · Preserved content.md, line 188 (Sevbot is built on top of Codex) ### Model unknown Inference · Catalog judgment · Medium confidence · `openai-sevbot--architecture-model` Confidence reason: The article names no underlying models; unknown does not mean absent. Evidence: - Supports · [1] [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/openai-sevbot-article/content.md) · The preserved article names no underlying models for Sevbot ### Interfaces slack Fact · Reported · Medium confidence · `openai-sevbot--architecture-interfaces` Confidence reason: A linked participant or independent source reports the claim. Evidence: - Supports · [1] [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/openai-sevbot-article/content.md) · Preserved content.md, line 192 (it is part of the Slack channel and answers devs' questions) ## Lessons and interpretation ### Lesson OpenAI's stated goal is that Sevbot mitigates routine outages autonomously with humans reviewing its actions on return; as of the report, on-call duty remains Fact · Reported · High confidence · `openai-sevbot--lessons-learned-0` Confidence reason: The stated goal and the current on-call state are reported directly. Qualifications: - Observation date: 2026-09 Evidence: - Supports · [1] [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/openai-sevbot-article/content.md) · Preserved content.md, line 195 (goal of autonomous action on routine outages; oncall duty is not a thing of the past) ## Related implementations - This entry is a component of: [OpenAI — Agentic software factory](https://internal-agents.com/agents/openai-software-factory) ## Sources 1. [Inside OpenAI's agentic software factory](https://newsletter.pragmaticengineer.com/p/openai-software-factory) - News · Independent secondary · Evidence - Original URL: - Publisher: The Pragmatic Engineer · Published: 2026-09-15 · Accessed: 2026-09-16 · Last verified: 2026-09-16 - Preserved copy in the repository: 2. [Certificate Transparency records for sevbot.api.openai.com](https://crt.sh/?q=sevbot.api.openai.com) - Other · Aggregator · Evidence - Original URL: - Publisher: crt.sh · Accessed: 2026-09-16 · Last verified: 2026-09-16 - Preserved copy in the repository: 3. [\[SEVBot\] Optionize initialized notification tolerance (openai-agents-python pull 2765 commits)](https://github.com/openai/openai-agents-python/pull/2765/commits) - Source code · First party · Evidence - Original URL: - Publisher: OpenAI · Published: 2026-03-24 · Accessed: 2026-09-16 · Last verified: 2026-09-16 - Preserved copy in the repository: