AgentMart

Store in my memory locker

locker_put

DestructiveIdempotent

Store a value in your memory locker under a key (survives between sessions). Send exactly one of value (text) or value_base64 (bytes). The answer holds the value's etag: pass it as if_match next time so that a run that overlaps yours cannot be overwritten without notice.

Input Schema

TableJSON Schema

Name	Required	Description
`key`	Yes	Key, 1 to 200 characters from A-Z a-z 0-9 . _ : / - (no spaces; "/" separates folders).
`value`	No	The value as text (UTF-8). "" stores an empty value.
`if_match`	No	Store only if the current value has this etag (from locker_get or locker_put); "*": only if the key exists. Otherwise 412 precondition_failed and nothing is stored.
`instance_id`	Yes	Utility instance id from buy or list_utilities.
`content_type`	No	Stored and returned on read. Default text/plain; charset=utf-8 for value, application/octet-stream for value_base64.
`value_base64`	No	The value as bytes, in standard base64 with = padding.
`if_none_match`	No	"*": store only if the key does not exist yet (create only), else 412 precondition_failed.
`passport_token`	No	Your amp_ token, only if your client cannot send it as an Authorization header; leave it out when your MCP app signed in.

Schema Changelog

Changes observed during successful MCP inspections.

Changed8 schema fields changedOct 5, 2026

addedInput schema / additionalProperties
```
Added value: +false
```

addedInput schema / properties / content_type / description

Added value: +"Stored and returned on read. Default text/plain; charset=utf-8 for value, application/octet-stream for value_base64."

addedInput schema / properties / if_match

Added value: +{
+  "description": "Store only if the current value has this etag (from locker_get or locker_put); \"*\": only if the key exists. Otherwise 412 precondition_failed and nothing is stored.",
+  "maxLength": 200,
+  "minLength": 1,
+  "pattern": "^[\\x20-\\x7e]+$",
+  "type": "string"
+}

addedInput schema / properties / if_none_match

Added value: +{
+  "const": "*",
+  "description": "\"*\": store only if the key does not exist yet (create only), else 412 precondition_failed.",
+  "type": "string"
+}

addedInput schema / properties / key / description

Added value: +"Key, 1 to 200 characters from A-Z a-z 0-9 . _ : / - (no spaces; \"/\" separates folders)."

changedInput schema / properties / passport_token / description

Previous value: -"Leave it out when your MCP app signed in to AgentMart (it is refused then). Otherwise your passport bearer token (amp_...) or session token (amp_s_...), only if your client cannot send it as an Authorization header. A token here sits in your context, so it never authorizes sensitive actions."New value: +"Your amp_ token, only if your client cannot send it as an Authorization header; leave it out when your MCP app signed in."

addedInput schema / properties / value / description

Added value: +"The value as text (UTF-8). \"\" stores an empty value."

addedInput schema / properties / value_base64 / description

Added value: +"The value as bytes, in standard base64 with = padding."

First observedOct 4, 2026

TDQS

A4.2/5.0

Behavior4/5

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

Annotations already declare destructive and idempotent hints, but the description adds real context beyond them: persistence across sessions, the overwrite-protection rationale, and the returned etag. It does not, however, spell out what an overwrite destroys or the exact 412 outcomes (those sit in the schema).

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?

Three tight sentences, front-loaded with purpose, then the mutual-exclusivity rule, then the concurrency pattern. Every sentence carries decision-relevant content with 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?

With 8 parameters at full schema coverage and no output schema, the description adequately compensates by naming the returned etag and the concurrency contract. It is close to complete, only omitting a fuller statement of what a write overwrites and the failure outcomes.

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, but the description adds a constraint the schema does not encode: value and value_base64 are mutually exclusive ('send exactly one of'). It also frames if_match as consuming the prior etag, giving the parameter operational meaning.

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 (store) and resource (memory locker) scoped by key, plus the key distinguishing trait that values survive between sessions. The verb alone cleanly separates it from locker_get, locker_delete, and locker_list 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 Guidelines3/5

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

The description gives parameter-level routing ('send exactly one of value or value_base64') and explains the etag/if_match concurrency pattern, but offers no explicit when-to-use versus siblings like locker_get or locker_delete, nor when to prefer a conditional write. Usage is implied rather than stated.

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

All AgentMart 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.

Resources

Need Help?