Skip to main content
Fleet runs on native dspy.RLM (DSPy 3.3.0b1). Every Turn constructs one fresh dspy.RLM, calls it once, and settles. There is no dspy.ReAct chat wrapper, no long-lived Agent, and no Fleet-owned RLM variable-mode monkeypatch. The maintained integration surface is the shipped Signatures, the delegation ladder, and the instruction fragments that compose the Root prompt.

​
Where DSPy lives

​
Configure models through profiles

Model, provider, token, and recursion policy live in config/fleet.toml under the selected profile. There are no DSPY_* environment variables. Select the profile at the top of config/fleet.toml:
Provider environment variables are set by profile:
  • Interactive profiles (daytona, daytona-recursive) call OpenCode Go: FLEET_OPENCODE_GO_API_KEY, FLEET_OPENCODE_GO_BASE_URL.
  • Managed and benchmark profiles call the Databricks AI Gateway: DATABRICKS_TOKEN, FLEET_DATABRICKS_AI_GATEWAY_BASE_URL.
  • All committed profiles use deepseek-v4-flash for both Root and Sub.
See the configuration reference for the full matrix.

​
The Root Signature

The default Fleet Root Signature carries a bounded, strict input surface. Every Signature receives:
  • request text.
  • Bounded session_context.
  • Bounded skill_cards.
  • Bounded Attachment metadata.
Full committed history stays host-side behind the read_session_history Tool. The Signature uses strict local Pydantic DTOs; conversion and JSON serialization happen once immediately before native dspy.RLM.acall(). Custom Skill Signatures retain JSON-compatible common input annotations. Only one selected Skill may provide a validated custom Signature per Turn; data-analysis is the only bundled Skill that does so.

​
Instructions are composed, not monolithic

src/fleet_rlm/rlm/instructions.py owns the default Fleet Root instruction fragments: base, REPL, tool, optional recursion, verification, and bounded-context guidance. Fragments are composed directly; disabling recursion under a non-recursive profile omits recursion guidance rather than deleting text from one large monolithic docstring.

​
Delegation ladder

The Root selects the cheapest sufficient primitive:
  1. Python in the interpreter for deterministic work.
  2. Native llm_query / llm_query_batched for semantic work.
  3. rlm_query for one iterative isolated subproblem.
  4. Root-only rlm_query_batched for ordered independent child RLMs.
Recursive children remain one native level deep. RLM_NATIVE_CHILD_DEPTH = 1 is a fixed product invariant, not a policy value. Fleet reserves the shared recursive budget atomically before starting a child and bounds sibling concurrency through recursion_max_parallel_children. See the Recursive RLM concept page for the isolation contract and the [rlm] bounds.

​
Large inputs

Long documents, workspace bundles, and durable Attachments cannot be inlined into the Root prompt. DSPy 3.3.0b1 ships the SandboxSerializable contract; Fleet builds host-constructed capsules that DSPy injects into the interpreter as REPL variables. For example, authorized Attachment context is packaged into AttachmentContextCapsule before the Turn runs:
Inside the sandbox, DSPy reconstructs the value through sandbox_assignment(...) while the LM sees only a short rlm_preview() summary. Skill authors should not import capsule classes directly; the host builds them.

​
Interpreter reuse within one Turn

Within one Run, interpreter calls reuse one context. Python state persists across RLM iterations. Every later Run receives a fresh context, and replacing a Daytona Sandbox remounts the Workspace Volume Scope without preserving Python globals.

​
Tracing

config/fleet.toml [mlflow] policy controls DSPy tracing. Fleet enables MLflow DSPy inference autologging for the selected experiment; compile and evaluator traces stay disabled for live Turn observability. mlflow.trace_content_max_chars bounds each readable field, and mlflow.async_logging = true keeps trace export off the Turn critical path. See the observability page for the full policy.

​
What is not here

  • No dspy.GEPA optimization API surface. Committed profiles are deterministic; there is no optimize subcommand and no POST /api/v1/optimization/* endpoints.
  • No dspy.ReAct FleetAgent chat wrapper.
  • No BYOK per-caller LM selection.
  • No Fleet-owned RLM variable-mode wrappers. Large inputs use DSPy’s native SandboxSerializable.

​
See also

Recursive RLM

One-level recursive child boundary and [rlm] bounds.

Agent model

The one-Turn, one-fresh-RLM model and Skill disclosure.

HTTP API

Turn SSE contract that every Signature runs behind.

Configuration

Profile matrix and provider environment variables.
Last modified on August 13, 2026