Skip to main content
Glama
Sign In

past.dev

by dev.past
OverviewDiscussionsAdmin

Ingest into a project's memory

ingest_memory
DestructiveIdempotent

Ingests one data point into a project's memory, as POST /api/v1/ingest does, and answers as it does: the ingestion id, the source id and the status. With no audience the data point is for the whole project; name an audience by its slug to keep it to that audience. An id that already exists replaces the earlier revision and the memories that came from it, and the same content under the same id is unchanged and costs nothing. It spends the project's credits as an API ingestion does. Ingest only what the person asks you to keep. Needs project Write.

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": {
      "content": "<content>"
    },
    "name": "ingest_memory"
  }
}

Input Schema

NameRequiredDescriptionDefault
idNoYour own id for this data point, with no '/' in it. Send it again to replace the earlier revision. When omitted, the id comes from the content.
labelNoA short title for the data point, such as "Email from Nadia". When omitted on a re-send, the stored label stays.
contentYesThe text to keep, as the person gave it.
projectNoThe project's slug, exactly as list_projects returned it. Omit it only when the organization has one live project.
audienceNoThe slug of an audience that exists in the project, to keep the data point to that audience. Omit it for the whole project.
identityNoThe author: the customer's own id for one of their users. Its first appearance creates it in the identity directory.
metadataNoYour own JSON, stored and returned as sent. Keys such as role, author, conversationId and messageId help past attribute and order what it reads.
timestampNoWhen the data point happened (ISO 8601). Now when omitted, or the stored time on a re-send.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • changedInput schema / properties / id / description
      Previous value: -"Your own id for this data point. Send it again to replace the earlier revision. When omitted, the id comes from the content."New value: +"Your own id for this data point, with no '/' in it. Send it again to replace the earlier revision. When omitted, the id comes from the content."
  2. Changed5 schema fields changed
    • changedInput schema / properties / audience / type
      Previous value: -"string"New value: +[
+  "string",
+  "null"
+]
    • changedInput schema / properties / id / type
      Previous value: -"string"New value: +[
+  "string",
+  "null"
+]
    • changedInput schema / properties / identity / type
      Previous value: -"string"New value: +[
+  "string",
+  "null"
+]
    • changedInput schema / properties / label / type
      Previous value: -"string"New value: +[
+  "string",
+  "null"
+]
    • changedInput schema / properties / project / type
      Previous value: -"string"New value: +[
+  "string",
+  "null"
+]
  3. Added

TDQS

A4.6/5.0
Behavior5/5

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

Goes well past the annotations: it discloses the exact return payload (ingestion id, source id, status), the revision/replacement behavior behind destructiveHint, why re-sending identical content under the same id is free (idempotentHint), that it consumes project credits, and that it requires project Write. That is precisely the context an agent needs before a mutating, credit-spending call.

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 the action and return shape, then scope rules, then cost and permission constraints — a good ordering with no filler. Slightly dense in the middle where the audience and id-replacement rules are packed into single sentences, but every sentence carries information.

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?

Covers action, return values, scope, idempotency, destructive replace semantics, credit cost, and authorization, which is everything needed for an 8-parameter mutation with no output schema. Nothing material is missing.

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 still adds the behavioral meaning of `audience` (project-wide vs audience-scoped) and of re-sending an existing `id` ('replaces the earlier revision and the memories that came from it'), which the schema only hints at. It does not touch identity, metadata or timestamp, but those are fully documented in 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?

States a specific verb and resource ('ingests one data point into a project's memory') and anchors it to a known equivalent endpoint ('as POST /api/v1/ingest does'), which separates it from recall_memory, estimate_ingestion and explain_ingestion. The replace/scope semantics immediately clarify what kind of write this is.

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 exclusion ('Ingest only what the person asks you to keep') and conditions the audience argument ('With no audience the data point is for the whole project; name an audience by its slug to keep it to that audience'). It does not name a sibling alternative such as estimate_ingestion for cost estimation or recall_memory for reading back, so it stops short of full routing guidance.

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

All past.dev toolsInspect
Try in Browser
Claude Code, inside the Godot 4 editorFlockbay puts an AI agent in the real Godot 4 editor: it writes GDScript, runs your game and fixes what breaks. Free, or bring your own Claude Code or Codex.flockbay.comAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources