SaibleDocs
Context

What each tool is wired to

Read from its own sources, confirmed by a person, signed. A change waits for a decision.

Readrepository, accounts,traces, MCP serversProposea signed manifesthashes in the clearDriftagainst theconfirmed oneConfirmowner or admin+ codeBwidening: an adminSigned bundlecarries its hashgateway: not used yetContext APIa signed receiptfor its agents↺ read again every night;a GitHub push or a Bedrock change starts one sooner
From a tool's sources to a confirmed manifest · read only, never written to
  • Repository
    prompts, tool schemas, MCP configs · GitHub push
  • Amazon Bedrock
    agents, prompts, knowledge bases, guardrails · EventBridge
  • Snowflake
    Cortex Search, Cortex Agents, column tags
  • OpenAI
    projects, models used, Assistants, vector stores
  • Databricks
    serving endpoints, AI Gateway, vector search
  • Your traces
    what the tool was seen to call, by hash
  • MCP servers
    Confluence, SharePoint, Salesforce, Google Drive, any
  • The registration
    purpose, team, approved models, what the team declares
Where context is read from · tick: takes change events · no reader has yet run against a live account
  • In the clearHashes, structure and signatures: which models, which prompts by hash, which tools by schema hash, which sources with their tags.
  • Kept encryptedA prompt's text and the full manifest: in your own bucket under your KMS key when you set one, otherwise in Saible's store under your organisation's key.
  • Never keptA change event's body: no commit message, no author, no request parameters. Only names and ids.
What Saible keeps of a tool's context
Do it
  1. 1An org admin connects a source with POST /api/v1/context/sources; the console has no form for it yet.
  2. 2On Context, under Connected systems (MCP), an org admin registers a server and clicks Attach to give it to a tool.
  3. 3On the tool's page, click Read now and read the proposed manifest against the confirmed one.
  4. 4Click Confirm v…: as the tool's owner after MFA, or as an org admin through the review queue.
  5. 5Open Ontology: every element, where it is read from, and which tools use it.
For your agents

This topic is also written as Agent Skills your agents can load: context-layer, context-events, mcp-connected-systems, business-ontology, ontology, ontology-sources, context-health, ontology-api. Ask us for them.

Next · Control Plane · EvaluationCandidate models, on your own prompts →