noticed

log_interaction

Record a touchpoint — say WHAT it was with type: 'in_person', 'call', or 'message'. Everything else is optional detail that defaults sensibly. Builds the relationship timeline and feeds recency and relationship strength. Each call appends a new event. Works on any search_people hit — if they're not in your network yet, they're added first. NOT for notes — a note about someone is a memory, so use add_memory instead.

Input Schema

TableJSON Schema

Name	Required	Description
`tags`	No	Event/context tags for THIS touchpoint — WHERE / WHEN you met: an event, a place, a trip (e.g. 'friendly-machines-2026', 'event:agents-day', 'berlin'). Put them here, on the interaction, not on the person — they describe the meeting, not a durable trait, and are surfaced back on the person view as aggregated context. Use update_person's tags only for lasting traits of the person.
`type`	Yes	What the interaction was: 'in_person' (you were physically together), 'call' (a live conversation — phone or video), or 'message' (an asynchronous written exchange).
`format`	No	Group size, for in_person and call only — 'one_to_one' (default) or 'group'. A group counts for less than a 1:1. A message has no size.
`channel`	No	Where it happened. For a call: 'phone' (default) or 'video'. For a message: 'email' \| 'linkedin' \| 'x' \| 'whatsapp' \| 'sms'. Not used for in_person.
`payload`	No	Optional freeform details (e.g. { topic: 'fundraising' }).
`direction`	No	Messages only — 'outbound' (you sent it) or 'inbound' (they sent it). Set it whenever you know; it is what lets a real back-and-forth be told apart from a message that was never answered. Omit it when you genuinely don't know: the message is still recorded and still counts, it just carries no direction verdict. Never guess. A call or an in-person meeting is a two-way event and carries no direction.
`person_id`	Yes	The person this interaction is with.
`occurred_at`	No	When it HAPPENED (ISO timestamp). Omit → defaults to now. Pass null for undated items (e.g. open-ended follow-ups) — the timeline then shows only when it was recorded.

Output Schema

TableJSON Schema

Name	Required	Description
`ok`	Yes	Whether noticed completed the operation.
`data`	No	The operation result when ok is true.
`error`	No	A human-readable error when ok is false.

Schema Changelog

Changes observed during successful MCP inspections. Dates show when Glama detected each change.

Changed1 schema field changedAug 22, 2026

changedOutput schema / (root)

Previous value: -nullNew value: +{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "additionalProperties": true,
+  "properties": {
+    "data": {
+      "additionalProperties": true,
+      "description": "The operation result when ok is true.",
+      "properties": {},
+      "type": "object"
+    },
+    "error": {
+      "description": "A human-readable error when ok is false.",
+      "type": "string"
+    },
+    "ok": {
+      "description": "Whether noticed completed the operation.",
+      "type": "boolean"
+    }
+  },
+  "required": [
+    "ok"
+  ],
+  "type": "object"
+}

First observedAug 22, 2026

TDQS

A4.7/5.0

Behavior5/5

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

Beyond the annotations, the description discloses that each call appends a new event (non-idempotent), that the tool builds the relationship timeline and affects recency/strength, and that people may be added to the network as a side effect. These behaviors are not visible in the annotations and are clearly surfaced.

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?

The description is six tight sentences with no filler: purpose, defaults, side effects, and exclusions each get exactly one clear sentence. The most important action ('Record a touchpoint') is front-loaded, and every sentence earns its place.

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

Completeness5/5

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

For a tool with a rich schema and an output schema, the description covers what an agent needs to select and invoke it correctly: required type values, optional defaults, side effects, network behavior, and the key sibling alternative. Nothing essential is missing.

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?

The schema already provides 100% parameter documentation, including enums and defaults, so the description does not need to repeat them. The description adds high-level guidance like 'Everything else is optional detail that defaults sensibly', but it does not materially enrich individual parameter semantics beyond the schema.

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

Purpose5/5

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

The description opens with a specific verb and resource ('Record a touchpoint') and enumerates exactly what the `type` field accepts. It also distinguishes itself from the closest sibling by saying 'NOT for notes — ... use add_memory instead', so an agent can select it without ambiguity.

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

Usage Guidelines5/5

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

It explicitly tells the agent when to use this tool ('Works on any search_people hit') and when not to use it ('NOT for notes'), naming the alternative tool. It also sets expectations about auto-adding people to the network, which is important contextual guidance for invocation.

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

All noticed tools Inspect

Try in Browser

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

TDQS

B3.4/5.0

Disambiguation3/5

The tool set is organized around distinct resources, and the descriptions work hard to separate them, but several close pairs remain easy to confuse: add_memory vs memory_save vs add_note, accept_identity_match vs suggest_identity_match, and dismiss_identity_match vs mark_different_people. An agent will often need to read very subtle signals (who originated the content, pending vs initiating a merge, soft vs durable rejection) to pick the right tool.

Naming Consistency3/5

Most tools follow a clear verb_noun snake_case pattern like create_list, update_person, and delete_view, which is readable and mostly predictable. However, the memory tools break the pattern (memory_save, memory_get, memory_search instead of save_memory/get_memory/search_memory), and a few noun-style names (my_profile, network_summary, account_status) add inconsistency.

Tool Count1/5

At 57 tools, this is an extremely large surface that exceeds the calibration threshold for an extreme mismatch. The scope is broad, but many tools are micro-specialized variations of the same concept, such as four memory-related tools and seven identity-match tools, which makes the count feel inflated rather than well-scoped.

Completeness4/5

The tool set provides thorough lifecycle coverage for the core domain: people can be added, updated, searched, and removed; lists, views, actions, and scheduled tasks have create/read/update/delete; and identity matching has accept, dismiss, differentiate, and suggest paths. Minor gaps exist, such as no direct memory/note deletion or intro deletion, but agents can generally complete workflows without hitting dead ends.

Resources

Need Help?