Featherlane AI
Concepts

Concepts

Architecture, policy model, agent profiles, and the plugin contract.

Featherlane AI is a runtime safety layer for production agents. The important moment is the boundary before an agent speaks, sends, writes, or takes another externally visible action.

Featherlane AI in one decision

Core Ideas

ConceptMeaning
CheckOne request asking whether a proposed output is safe to deliver.
Authorization decisionThe typed response from Featherlane AI. It includes the effect, reason, findings, latency, trace id, and durable authority references when applicable.
Authorization effectOne of permit, deny, transform, require_approval, or defer. Approval can satisfy only require_approval; it cannot bypass deny or defer.
PolicyA YAML rule that scopes when it applies, what it matches, and which action to take.
Agent profileContext about the agent's domain, authority, tone, and constraints.
TraceThe persisted record that lets operators inspect what happened later.

Approval And Execution Flow

TermPlain-language meaning
IntentThe durable record of the action the agent is proposing. It is not permission to execute.
ApprovalA pending human decision. Approving creates authority but does not execute the action.
GrantRevocable authority for one exact action or a bounded reusable scope. It never overrides current policy.
Authorization claimThe grant and attempt identifiers the caller presents when retrying the action.
LeaseA one-attempt execution right issued only after Featherlane AI rechecks current policy and live state.
Authorization receiptAudit evidence explaining the decision and linking its intent, approval, grant, and lease. It carries no execution rights.

The complete flow is:

  1. The agent proposes an executable action and Featherlane AI records its intent.
  2. Policies and live checks evaluate the action.
  3. If authority is missing, one domain-labeled item appears in Approvals.
  4. Approval creates an exact-once or bounded scoped grant.
  5. The caller retries with that grant; Featherlane AI rechecks current policy and live state.
  6. A one-attempt lease prevents duplicate execution.
  7. Featherlane AI writes an authorization receipt, while the domain records its own result.

Financial actions use this same approval flow. Approvals is the only place a person approves or denies them; Financial actions remains the ledger and execution-history view.

Runtime Shape

The Rust backend is the source of truth. SDKs call the Rust server directly for runtime decisions, especially POST /v1/events. The web dashboard may proxy and display data, but it does not own policy evaluation or trace storage.

Runtime data flow

Latency Model

Featherlane AI is designed around channel budgets:

ChannelTarget
Streaming chatdeterministic checks under the realtime budget
Chatsynchronous guardrail decisions before the message is shown
Email and jobsslower checks are acceptable because delivery is asynchronous
Replayoffline analysis can spend more time for richer evaluation

The hot path starts with deterministic policies because that is the only layer that can be consistently fast enough for realtime streaming.

Source Of Truth

The deeper concept docs live in docs/concept/ in the repo:

  • architecture.md explains request flow and ownership.
  • crates.md explains each Rust crate and dependency boundary.
  • glossary.md defines domain terms once.
  • plugin-contract.md describes extension boundaries.
  • v0-design-decisions.md records why the current runtime shape exists.

Diagram source lives in docs/diagrams/*.d2; run pnpm docs:diagrams to regenerate the SVGs used by this site and the repo Markdown docs.

On this page