Skip to main content
Glama
aethis-mcp

aethis-mcp

Official
by Aethis-ai

aethis_refine

Refine an existing published ruleset: add optional feedback, make the minimal edit to fix failing tests while keeping passing tests green, then re-run the full suite.

Instructions

Refine an existing published ruleset: add optional feedback, then make the MINIMAL edit to fix failing test cases while keeping passing tests green, and re-run the full suite (seed-from-existing incremental re-authoring). Use this to fix a specific finding without re-authoring the whole section; use aethis_generate_and_test for a from-scratch rebuild.

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": {
      "project_id": "<project_id>"
    },
    "name": "aethis_refine"
  }
}

Input Schema

NameRequiredDescriptionDefault
modelNoOptional authoring model
feedbackNoOptional correction or domain knowledge to add before regenerating
thinkingNoPer-generation thinking: disabled, adaptive, or enabled:N. Omit or null to inherit.
openai_keyNoRetired and refused: OpenAI credentials are not accepted for authoring.
project_idYesThe project ID
anthropic_keyNoAn Anthropic API key the user explicitly provided for this call. [sensitive — do not echo or log] Deprecated: the raw value is written verbatim to the host's session transcript. Never fill this from the environment.
deepseek_key_envNoOptional. For deepseek-flash only: must equal the variable explicitly configured in AETHIS_DEEPSEEK_KEY_ENV. Ambient provider keys are never read.
anthropic_key_envNoOptional. Only honoured when it equals the env var the user configured via AETHIS_ANTHROPIC_KEY_ENV in this MCP server's config; that configured key is used automatically, so this can be omitted. Do not guess a variable name: the server refuses any name the user did not configure.
anthropic_key_keychainNomacOS keychain reference the user created for Aethis: either 'service:account' or just 'account' (service defaults to 'aethis-anthropic-key'). The server reads it via the `security` command at call time.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed4 schema fields changedv0.23.0
    • addedInput schema / properties / deepseek_key_env
      Added value: +{
      +  "description": "Optional. For deepseek-flash only: must equal the variable explicitly configured in AETHIS_DEEPSEEK_KEY_ENV. Ambient provider keys are never read.",
      +  "type": "string"
      +}
    • addedInput schema / properties / model
      Added value: +{
      +  "description": "Optional authoring model",
      +  "enum": [
      +    "claude-sonnet-5",
      +    "deepseek-flash"
      +  ],
      +  "type": "string"
      +}
    • changedInput schema / properties / openai_key / description
      Previous value: -"Retired and refused: Aethis LLM tools use Anthropic models only."New value: +"Retired and refused: OpenAI credentials are not accepted for authoring."
    • addedInput schema / properties / thinking
      Added value: +{
      +  "description": "Per-generation thinking: disabled, adaptive, or enabled:N. Omit or null to inherit.",
      +  "type": [
      +    "string",
      +    "null"
      +  ]
      +}
  2. First observedv0.22.0

TDQS

A3.6/5.0
Behavior3/5

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

Adds genuine behavior beyond the annotations: edits are deliberately minimal, passing tests must stay green, and the full suite is re-run — effectively describing a test-driven correction loop. However, it does not warn that this is non-idempotent and open-world (per annotations), nor that it triggers a generation call with possible wait/status polling, nor that the sensitive key parameters are refused or deprecated.

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

Conciseness4/5

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

The core action is front-loaded in the first clause, and the second sentence handles routing. It is a single dense sentence with a parenthetical label, which is slightly packed but contains no filler.

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

Completeness4/5

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

For a 9-parameter mutation tool with full schema coverage and no output schema, the description covers what it does and when to use it adequately. The main residual gap is that it says nothing about what the call returns or whether the result surfaces as an in-progress generation, which matters for a test-running re-authoring operation.

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

Parameters3/5

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

Schema description coverage is 100%, so the schema already documents every parameter, including the sensitive-key and deprecated openai_key fields. The description only echoes the optional 'feedback' parameter, adding no new syntax, format, or precedence meaning beyond what the schema provides.

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?

States a specific verb and resource ('Refine an existing published ruleset') plus the actual mechanics (add feedback, minimal edit to fix failing tests, re-run full suite). It clearly distinguishes itself from aethis_generate_and_test, but leaves aethis_refine_fields and aethis_refine_sections unaddressed, so an agent cannot fully disambiguate the 'refine' family from the description alone.

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

Usage Guidelines4/5

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

Gives an explicit use case ('fix a specific finding without re-authoring the whole section') and names the alternative for the opposite case ('use aethis_generate_and_test for a from-scratch rebuild'). Missing guidance on when to prefer the sibling refine_fields/refine_sections tools, which is the most likely routing confusion.

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