Source: https://internal-agents.com/agents/domu-clementino # Domu — Clementino A general-purpose 'AI colleague' spanning sales, finance, client ops, engineering, and recruitment, later split into a reusable toolkit plus Slack and desktop surfaces. - Approach type: Task agent - Deployment stage: Deployed - Autonomy: Human in loop - Evidence strength: Detailed primary - Status: Internal - First reported year: 2026 - Work: Support, Finance ops, Coding, Recruitment, Customer success - Interfaces: Slack, Desktop - Invocation: Interactive - Entry reviewed: 2026-09-09 ## Where people stay involved - **employee request → approved customer-impacting action** — Work product review · Level 3 ## Overview ### Summary A general-purpose 'AI colleague' spanning sales, finance, client ops, engineering, and recruitment, later split into a reusable toolkit plus Slack and desktop surfaces. Fact · Reported · High confidence · `domu-clementino--summary` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ## How it works ### Supporting component Transient conversation, persistent facts, knowledge RAG, and live system state; kept separate Fact · Reported · High confidence · `domu-clementino--primitives-0` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Supporting component Capabilities separated from the interface layer so they power multiple surfaces Fact · Reported · High confidence · `domu-clementino--primitives-1` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ## Supervision evidence ### Operating model assessment Level 3 for employee request → approved customer-impacting action; human attention boundary: work-product-review. Inference · Catalog judgment · High confidence · `domu-clementino--operating-models-0` Confidence reason: The source documents explicit human approval for customer-impacting actions. Qualifications: - Observation date: 2026 Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ## Implementation details ### Harness Claude/Anthropic SDK wrapper; a reusable tool/skill/memory layer separated from the Slack and desktop interfaces Fact · Reported · High confidence · `domu-clementino--architecture-harness` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Model Claude (Anthropic SDK) Fact · Reported · High confidence · `domu-clementino--architecture-model` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Interfaces slack, desktop Fact · Reported · High confidence · `domu-clementino--architecture-interfaces` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Tool access ~35 integrations organized into skills; specialist delegates per domain Fact · Reported · High confidence · `domu-clementino--architecture-tool-access` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Knowledge Four memory layers: conversation context, persistent facts, knowledge RAG, and live system state Fact · Reported · High confidence · `domu-clementino--architecture-knowledge` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Credentials Customer-impacting actions gated behind team-visible Slack approvals Fact · Reported · High confidence · `domu-clementino--architecture-credentials` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Context management Four-layer memory separation; prompt size dropped substantially after splitting capabilities from the Slack layer Fact · Reported · High confidence · `domu-clementino--architecture-context-mgmt` Confidence reason: A linked first-party source states the claim. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ## Reported metrics ### Headline claim ~35 integrations organized into skills Metric · Reported · High confidence · `domu-clementino--headline-metric` Confidence reason: A linked first-party source states the claim. Qualifications: - The source does not report the denominator of this figure. - Reported by: Domu - Scope: Clementino toolkit integration modules organized into skills Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) · Preserved content.md, lines 58–60 ### Key observation ~35 integrations organized into skills Metric · Reported · High confidence · `domu-clementino--key-metrics-0` Confidence reason: A linked first-party source states the claim. Qualifications: - The source does not report the denominator of this figure. - Reported by: Domu - Scope: Clementino toolkit integration modules organized into skills Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) · Preserved content.md, lines 58–60 ## Lessons and interpretation ### Lesson Separate reusable capabilities (tools/skills/memory) from the interface layer; prompt size drops while capability is preserved Inference · Catalog judgment · Medium confidence · `domu-clementino--lessons-learned-0` Confidence reason: The catalog derives this observation from the linked sources. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Lesson Model memory explicitly: conversation context, persistent facts, RAG, and live state have different lifecycles Inference · Catalog judgment · Medium confidence · `domu-clementino--lessons-learned-1` Confidence reason: The catalog derives this observation from the linked sources. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ### Lesson Gate customer-impacting actions behind human approvals rather than trusting the model Inference · Catalog judgment · Medium confidence · `domu-clementino--lessons-learned-2` Confidence reason: The catalog derives this observation from the linked sources. Evidence: - Supports · [1] [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) · [Preserved copy](https://github.com/steel-experiments/internal-agents-map/blob/main/archive/sources/domu-clementino-source-1/content.md) ## Sources 1. [Why we split our internal agent in two](https://domu.ai/blog/why-we-split-our-internal-agent-in-two) - Engineering blog · First party · Evidence - Original URL: - Accessed: 2026-08-12 · Last verified: 2026-08-31 - Preserved copy in the repository: