Skip to main content
Glama
reflex-brain

reflex-brain

by dokiromboide

reflex_apply_refinement

Apply a stored refinement proposal exactly once, requiring explicit approval or auto-apply when the evidence threshold is met. Use it to commit reviewed agent improvements.

Instructions

Apply a stored proposal once. Explicit approval requires approved=true; auto_apply only accepts evidence above threshold.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "refinement_id": "<refinement_id>"
    },
    "name": "reflex_apply_refinement"
  }
}

Input Schema

NameRequiredDescriptionDefault
approvedNoExplicit caller approval of the reviewed proposal
auto_applyNoAuto-apply if evidence threshold met
refinement_idYesproposal_id returned by propose (legacy parameter name)

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.2.0

TDQS

A3.6/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description must carry the behavioral burden. It does disclose two meaningful traits: application happens exactly once, and auto_apply only fires when evidence clears a threshold, so a silent no-op is possible. It omits what state is mutated, whether the change is reversible, and what happens on failure or threshold miss — important gaps for a mutation tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two short sentences, front-loaded with the core action, then the mode constraints. Every clause carries information not already stated elsewhere in the sentence, and there is no preamble or filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Parameters are fully documented and the invocation modes are covered, which is adequate for calling the tool. But for a mutation with no annotations and no output schema, the description should say what applying actually changes in the system and what the caller gets back; neither is addressed.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3; the description earns one step above by explaining the relationship between the flags — approved=true is the explicit-approval path, while auto_apply is evidence-gated rather than unconditional. It still does not address what happens if both flags are true or false, leaving the precedence unresolved.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

"Apply a stored proposal once" gives a specific verb (apply) and resource (a stored proposal), and the qualifier "once" distinguishes it from repeatable or bulk operations. It does not name a sibling such as reflex_rollback as the inverse operation, so an agent must infer the boundary from the names alone.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description explains the two activation modes (approved=true vs auto_apply) and their gating conditions, which is real guidance for choosing parameters. It never states when to call this instead of reflex_get_refinement (to inspect first) or reflex_rollback (to undo), leaving the decision implied by the sibling names.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.