Calera Infinite Context (ICX) — Persistent Memory & Associative Recall

Verbatim Register Quoting

icx_quote_slot

Read-only

Retrieves exact character-for-character verbatim text and cryptographic SHA-256 hashes from document registers by document family and 1-based slot index without AI paraphrasing. Optional start (0-based character offset) and length (character count) return a verbatim slice of the stored slot text. The slice is the exact stored characters, not a summary. Omit both start and length to return the whole slot. If start or length is out of range, the call errors and does not clamp. sha256 and parent_sha256 are the SHA-256 of the entire stored slot. slice_sha256 is the SHA-256 of the characters returned in exact_quote. Use this tool when exact contractual wording, statutory clauses, license keys, or immutable code revisions must be quoted verbatim without AI modification. Do NOT use for semantic question answering (use icx_recall_scoped), and do NOT use for keyword exploration (use icx_search_facts). Read-only operation with zero side effects. Requires X-License-Key authentication.

Input Schema

Name	Required	Description
`index`	Yes	1-based ordinal member index (e.g. 1 for first revision, 2 for second revision). Minimum: 1.
`start`	No	Optional 0-based character offset into the stored slot. Omit together with length to return the whole slot. Out of range is an error. The returned text is a verbatim slice of the stored characters, not a summary.
`family`	Yes	Document family identifier specified during ingestion (e.g. 'contract.indemnity', 'msa.pricing', 'auth.tokens').
`length`	No	Optional character count beginning at start. Omit together with start to return the whole slot. A length that runs past the end of the stored text is an error. The returned text is a verbatim slice, not a summary.
`space_id`	No	Optional memory space partition where the family was stored. Defaults to caller default space.

Output Schema

Name	Required	Description
`index`	Yes	description=Ordinal slot index (1-based)
`family`	Yes	description=Document family category
`sha256`	Yes	description=SHA-256 of the entire stored slot. Document seal. Not the slice hash.
`status`	Yes	description=Retrieval status (VERIFIED_VERBATIM_QUOTE \| REGISTER_SLOT_EMPTY)
`space_id`	Yes	description=Target memory space partition
`source_uri`	Yes	description=Original document source URI
`exact_quote`	Yes	description=Verbatim stored characters returned. The whole slot, or the requested start/length slice. Not a summary.
`slice_start`	Yes	description=0-based character offset of exact_quote inside the stored slot. 0 when the whole slot is returned.
`slice_length`	Yes	description=Character count of exact_quote. Equals the stored slot length when start and length are omitted.
`slice_sha256`	Yes	description=SHA-256 of the verbatim characters returned in exact_quote.
`parent_sha256`	Yes	description=SHA-256 of the entire stored slot. Same value as sha256. This is the document seal.

Schema Changelog

Changes observed during successful MCP inspections.

First observedOct 26, 2026

TDQS

A4.4/5.0

Behavior3/5

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

Annotations only declare readOnlyHint=true, and the description reinforces this ('read-only operation with zero side effects') while adding genuinely useful behavior: out-of-range start/length errors rather than clamping, and the three distinct hash scopes (sha256/parent_sha256 over the whole slot vs slice_sha256 over returned characters). It also discloses the X-License-Key auth requirement. These are real additions beyond the annotation, but nothing is said about pagination limits, maximum slice size, or error codes, so it stops short of full behavioral disclosure.

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 core action and scope, then slicing semantics, then hash semantics, then routing guidance. Every sentence carries information, though 'The slice is the exact stored characters, not a summary' restates schema text and adds mild redundancy.

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?

An output schema exists, so return values need not be enumerated, yet the description still explains what each hash field represents. Auth requirements, error behavior, and sibling routing are all present, leaving nothing an agent needs in order to call this correctly.

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, and the description earns an increment by clarifying the start/length interaction (omit both for the whole slot, supply both for a slice) and by disambiguating what the returned hashes cover. It does not add format or unit details for space_id beyond the schema, keeping it below a 5.

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 (retrieves) with precise resources: exact character-for-character verbatim text plus SHA-256 hashes from document registers, addressed by family and 1-based slot index. The 'without AI paraphrasing' qualifier and the family/index addressing make it unmistakable against every sibling.

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?

Explicitly names the conditions that select this tool (exact contractual wording, statutory clauses, license keys, immutable code revisions) and names two alternatives with their conditions: icx_recall_scoped for semantic QA, icx_search_facts for keyword exploration. Both when-to-use and when-not-to-use are covered.

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

All Calera Infinite Context (ICX) — Persistent Memory & Associative Recall 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.