Skip to main content
Glama

NASC: Networked AI Session Continuity

nasc_preferences

The project's now register, portable working-style preferences, and per-platform vendor privacy posture. Focus is set at nasc_start or mid-session through update, which writes the register; next and open_questions remain narrative status fields and never create board items. view reads the raw register rows. get_prefs reads preferences; set_prefs stores one as active immediately (identical text is a no-op; near-duplicates are refused unless replace or force); compile_rules renders the managed block for a host. Scribe and the Control Panel deliver managed blocks, not the model, and a preference change arrives as a notification. get_vendor_policy returns the per-platform privacy posture the web-capture extension enforces; report_vendor_policy records content-free enforcement outcomes.

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": {
      "action": "<action>"
    },
    "name": "nasc_preferences"
  }
}

Input Schema

NameRequiredDescriptionDefault
whoNoOperator name label. Attribution comes from the authenticated account.
hostNocompile_rules target host (claude_code|gemini_cli|antigravity|codex|cline|cursor|claude_ai|chatgpt|grok_web|perplexity|agents_md). set_prefs takes target_mode and target_slugs; host is the legacy single target.
nextNoaction='update' only. Replaces next[].
whatNoaction='update' only. One-line note of what was just edited.
brainNoOptional target org (brain uuid). Ignored unless this account can already access it; it never grants access.
forceNoset_prefs custom: bypasses the near-duplicate guard.
scopeNoDefault person.
actionYes
layersNocompile_rules: personal compiles only your personal preferences, without the org layer. Default full.
paramsNoData-only parameters.
confirmNoaction='update' only. Lets pick_up overtake an item in flight under another person's live session.
pick_upNoaction='update' only. Open-item keys (decision:<id>) claimed on this session. A claim in flight under another person's live session returns as in_flight_warning, unwritten, unless confirm=true; a parked or released claim transfers and best-effort pings the previous holder.
projectNoProject slug.
replaceNoset_prefs custom: uuid of a live custom preference in the same scope, owner, project and host that this write supersedes.
blockersNoaction='update' only. Replaces blockers[].
done_addNoaction='update' only. Appends one item to done[].
outcomesNoreport_vendor_policy: content-free per-control enforcement outcomes.
priorityNoHigher wins ties. Default 100.
put_downNoaction='update' only. Keys this session carries to park; the person keeps the claim. Keys not carried return under not_carried.
platformsNoget_vendor_policy: optional platform slugs (claude_ai|chatgpt|grok|perplexity|github_copilot); default all.
pref_textNointent_key=custom only.
intent_keyNoCatalog intent; custom is instruction-only.
session_idNoOptional session id from nasc_start. Pins this call's authorization, org and project to that session.
target_modeNoset_prefs audience: every target, only target_slugs, or all except target_slugs.
target_slugsNoset_prefs exact surface slugs; non-empty for only/except, empty for all.
current_focusNoaction='update' only. Replaces this project's focus.
open_questionsNoaction='update' only. Replaces open_questions[]; these remain narrative status and do not create Open Items.
continuity_tokenNoOptional host conversation id. Resolves the active session and pins this call's authorization, org and project to it.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A3.9/5.0
Behavior4/5

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

With no annotations, the description carries the full burden and does substantial work: set_prefs is a no-op on identical text, refuses near-duplicates unless replace/force, and applies 'active immediately'; a pick_up claim in flight returns in_flight_warning unwritten unless confirm=true, while parked claims transfer with a best-effort ping. It also discloses that Scribe and the Control Panel (not the model) deliver managed blocks and that preference changes arrive as notifications. Missing only response/error shape detail.

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?

Front-loaded with a resource summary before the action-by-action breakdown, and nearly every sentence carries a distinct rule (no-op, near-duplicate refusal, in_flight_warning, compile host). Dense but not padded; the main cost is that seven actions are compressed into a single paragraph, making scanning harder.

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 28-parameter, seven-action tool with nested objects, no annotations and no output schema, the description is unusually complete: it explains register mutation rules, preference lifecycle, claim/confirm mechanics, and managed-block delivery. Gaps are limited to return-value detail (e.g., what view actually yields) and the report_vendor_policy outcomes contract, which the schema partly covers.

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 96%, so the baseline is 3. The description reinforces a few bindings (next/open_questions as narrative status, host as legacy single target, confirm/replace/force semantics), but adds little new meaning beyond what the already-rich schema states.

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?

The opening sentence names three distinct resources (project now register, portable preferences, per-platform vendor privacy posture) and then maps each action verb to its resource ('view reads the raw register rows', 'get_prefs reads preferences', 'compile_rules renders the managed block'). Sibling differentiation is partial: it distinguishes next/open_questions from board items and pins focus-setting to nasc_start, but never names a sibling tool to route to for the non-preference cases.

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?

Per-action routing is explicit ('Focus is set at nasc_start or mid-session through update', 'view reads the raw register rows. get_prefs reads preferences'). It conveys context for each action, but gives no explicit when-not or hand-off to alternatives for the tool as a whole (e.g., when to use nasc_decisions instead).

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources