Wever Labs Agent Products

Memory retention check

wever_memory-retention-check

Read-only

Agent Memory: submit a memory setup’s retention: the stated period (a number of days, weeks, months or years, or forever), what it counts from (when a memory was written, when it was last used, or when the relationship ended), whether it is written where the principal can read it, how it is applied (old memories purged automatically, reviewed on a schedule with how often, cleaned up by hand, or not at all) and what happens to memory when the relationship ends (deleted with how soon, kept anonymized, or kept), plus optionally the categories with a period of their own, the stored memories seen with when each was written and last used, and the date they are read at. The signed report says whether a stated retention period exists, whether old memories are reviewed or purged on it and what survives the end of the relationship, and names no period, a period of forever, a period nobody applies, memory kept after the relationship ends, every category kept forever and every memory seen past its period, then a period not written down, cleanup only by hand, a review with no stated schedule and an anonymized copy surviving the end, with the fix for each. This reads how long the agent keeps what it remembers about its principals: the retention of records and data generally in a written policy is wever_data-retention-check, whether one removal request clears memory everywhere is wever_memory-deletion-check and what the agent keeps at all is wever_memory-scope-check. Verdict kept_forever, retention_gaps, retention_unverified or retention_enforced. POST /api/memory-retention-check. Reads only the material in the call: no network request, fetch, money movement, message, sharing, configuration change or external write. Paid URL: 0.10 USDC per call on Base via x402 exact, with the shared 10 signed-wallet free calls per rolling 30 days. Bind X-Wever-Wallet-Proof to the exact JSON arguments and the tool route. Returns a report with an Ed25519 signed receipt. Contract: /api/agent-memory-contract. Human page: https://weverlabs.com/agent-memory/.

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": {
      "retention": "<retention>"
    },
    "name": "wever_memory-retention-check"
  }
}

Input Schema

Name	Required	Description
`as_of`	No	The date the memories are read at, such as today.
`label`	No	Optional name recorded in the report.
`system`	No	The agent, such as the travel assistant.
`memories`	No	Stored memories seen, with their dates, to find any kept past their period.
`retention`	Yes
`categories`	No	Categories with a retention period of their own.

Schema Changelog

Changes observed during successful MCP inspections.

AddedNov 10, 2026

TDQS

A4.6/5.0

Behavior5/5

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

Annotations already declare readOnlyHint=true, destructiveHint=false and openWorldHint=false, and the description goes well beyond them: it states that only the material in the call is read, no network request, fetch, money movement, message, sharing, configuration change or external write occurs, and it discloses the paid path (0.10 USDC per call on Base via x402 exact, 10 signed-wallet free calls per rolling 30 days) plus the X-Wever-Wallet-Proof binding requirement and the Ed25519 signed receipt. Those are exactly the operational facts an agent needs before calling a paid, proof-gated tool.

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

Conciseness3/5

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

The content is front-loaded and every section is on-topic, but the prose is heavily run-on, with long comma-chained sentences packing field enumerations and a second dense sentence listing every verdict condition. Some material, such as the full enumeration of gap conditions alongside the verdict names, is close to redundant with the verdict list and could be tightened without losing meaning.

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 nested six-parameter tool with no output schema, the definition covers inputs, what the report asserts, the four possible verdicts, cost and auth, and the contract/page references, which is close to sufficient. What is left thin is the report's shape beyond its verdict (what a signed receipt contains and how the fix text is presented), so an agent knows the outcome labels but not the exact response structure.

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 already high at 83%, and the description still adds interpretive meaning by explaining the semantics of each retention facet: what the period counts from (written, last used, relationship end), how it is applied (automatic purge, scheduled review with frequency, manual, none), and what survives the end (deleted with how soon, anonymized, kept). It also explains why stored memories with written/last-used dates are supplied (to find memories kept past their period) and that the read date anchors the evaluation, which is more than the schema labels alone convey.

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 concrete verb and resource: submit a memory setup's retention and get back a signed report. It states exactly which facts are evaluated (period, counts-from, written-down, application, end-of-relationship handling) and names the three sibling checks it is not (wever_data-retention-check, wever_memory-deletion-check, wever_memory-scope-check), so an agent can distinguish it without opening any schema.

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?

Usage is explicitly routed: general records/data retention policy goes to wever_data-retention-check, whether one removal request clears memory everywhere goes to wever_memory-deletion-check, and what the agent keeps at all goes to wever_memory-scope-check. The scope sentence ('reads how long the agent keeps what it remembers about its principals') sets the boundary clearly enough to choose correctly.

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

All Wever Labs Agent Products tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.