Skip to main content
Glama

DOC Intel

doc_intel
DestructiveIdempotent

Analyze readable UTF-8 text or manage a private, resumable file-analysis job. Public text URLs are bounded to 1 MiB; pasted text to 60,000 UTF-16 code units and the complete MCP JSON request to 64 KiB. Private files require an active Zambo capability key and stay in private App Storage; use job_action=create, upload_url, upload_complete, part, status, pause, resume, cancel, review, delete or list. Files are limited to 25 MiB; PDFs to 200 pages; ZIPs to 20 MiB compressed, 50 MiB expanded, 100 text entries and 100:1 expansion. PDFs, images, archives and private repositories are not executed. Scanned pages/images require explicit Groq OCR consent, billed to Zambo with no separate user charge; exact provider USD cost is unavailable. Each analysis part covers at most 24,000 UTF-16 code units; file jobs resume after reconnect, lock source revisions, do not advance failed parts, and require complete range reconciliation and owner review. For larger text jobs keep the same _session_id, continue with offset and expected_source_sha256, collect every range and review before synthesizing. Groq fallbacks use the existing provider account, not a paid overflow service. When unavailable, YOU (the connected AI) can analyze the authenticated source range and submit host_analysis with host_handoff_token using this same tool. A pending handover is not completion and spends no free call; submission is labeled host-AI reported, never independently verified. Use analysis_mode=host_ai to skip Zambo inference entirely. Your AI's own plan still applies. Set require_semantic_analysis:true to retain an error until semantic analysis is submitted. doc_intel includes one completed first-range document preview per network per UTC day. Continuing analysis at next_offset and private file jobs require an active Day Pass or Zambo Pass.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "doc_intel"
  }
}

Input Schema

NameRequiredDescriptionDefault
urlNoPublic HTTPS UTF-8 text or HTML URL, up to 1 MiB. Not a binary PDF, archive, image or authenticated repository.
_refNoOptional handover referral ID from your install link.
noteNoOptional owner review note.
sizeNoFile size in bytes; file bytes upload directly to the private upload URL.
textNoExtracted readable text. Full JSON must fit 64 KiB; use smaller excerpts for multi-byte text. URL takes precedence if both are supplied.
emailNoLegacy compatibility field; email alone never establishes paid access. Use the connected pass credential.
job_idNoPrivate job ID returned by create.
offsetNoInitially 0. Continue with the exact coverage.next_offset returned by the previous part.
decisionNoOwner review decision after all ranges have been reconciled.
filenameNoOriginal filename for a private file job.
allow_ocrNoExplicitly permit sending only scanned pages/image files to Groq for OCR after reviewing the returned provider cost disclosure.
_zambo_keyNoOptional opaque Zambo access capability. Never place it in a URL or public log.
job_actionNoOptional private-file job action. Omit to use bounded text/URL analysis.
_session_idNoOptional stable bot-job ID. Reuse it across calls to bind a Day Pass and keep the job ledger together.
content_typeNoDeclared MIME type, checked again against extracted content.
_caller_labelNoOptional caller label recorded as a caller-reported claim in the receipt.zambo-starter
analysis_modeNoauto tries Groq then delegates to the connected AI. host_ai sends no Zambo inference requests; analyze the returned source excerpt yourself, then submit with this same tool.
host_ai_modelNoYour actual model name, or 'not disclosed'. Stored as caller-reported provenance.
host_analysisNoYour structured analysis of the recorded range: required title, summary and key_points array; optional doc_type, entities, risks, action_items, sentiment, confidence, word_count_estimate, brief. Recorded as caller reported, not independent verification.
_caller_handleNoOpaque anonymous continuity handle returned during initialize. Reuse it across MCP sessions; it is not payment authorization.
host_handoff_tokenNoPrivate recorded range token from host_handoff.token. Never share it in a public receipt or link.
receipt_visibilityNoReceipt visibility for this call. Public is the default. Private returns a bearer link; anyone with that link can read it, so keep it secret. The raw token is not stored.public
_continuity_ai_labelNoOptional self-reported AI client label for the private continuity timeline. Zambo does not independently verify the client.
expected_source_sha256NoPrevious coverage.source_sha256; required for offset > 0. Stops mixed-revision analyses.
require_semantic_analysisNoIf true, retain an error until semantic analysis is complete. A recorded host-AI handover can accompany the error; it is not completed analysis.

Output Schema

NameRequiredDescriptionDefault
okYes
idsYes
nextNo
tierNo
urlsYes
errorNo
graphNo
activeNo
resultYesThe tool's observed JSON value or plain-text result.
intentsNo
next_upNo
paymentNo
day_passNo
today_urlNo
expires_atNo
human_lineYes
next_stepsNo
daily_recapNo
renew_nudgeNo
unlock_linkNo
seconds_leftNo
resume_statusNo
resumed_resultNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed2 schema fields changed
    • addedInput schema / properties / receipt_visibility
      Added value: +{
      +  "default": "public",
      +  "description": "Receipt visibility for this call. Public is the default. Private returns a bearer link; anyone with that link can read it, so keep it secret. The raw token is not stored.",
      +  "enum": [
      +    "public",
      +    "private"
      +  ],
      +  "type": "string"
      +}
    • changedOutput schema / required
      Previous value: -[
      -  "result",
      -  "ok",
      -  "human_line",
      -  "next",
      -  "ids",
      -  "urls"
      -]New value: +[
      +  "result",
      +  "ok",
      +  "human_line",
      +  "ids",
      +  "urls"
      +]
  2. Changed1 schema field changed
    • changedInput schema / properties / require_semantic_analysis / description
      Previous value: -"If true, retain an error until semantic analysis is complete. A recorded host-AI handoff can accompany the error; it is not completed analysis."New value: +"If true, retain an error until semantic analysis is complete. A recorded host-AI handover can accompany the error; it is not completed analysis."
  3. Added
  4. Removed
  5. Added
  6. Removed
  7. Changed1 schema field changed
    • changedOutput schema / (root)
      Previous value: -nullNew value: +{
      +  "additionalProperties": true,
      +  "properties": {
      +    "result": {
      +      "type": "string"
      +    }
      +  },
      +  "type": "object"
      +}
  8. Added

TDQS

A4.2/5.0
Behavior5/5

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

Annotations already flag destructiveHint=true, but the description goes well beyond them: it discloses billing behavior (OCR billed to Zambo, exact provider USD cost unavailable), quota ('one completed first-range document preview per network per UTC day'), job semantics (resume after reconnect, lock source revisions, do not advance failed parts), and provenance limits (submissions labeled host-AI reported, never independently verified). This is the kind of consequence-and-cost context annotations cannot carry.

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?

Purpose is front-loaded and every sentence carries facts, but the prose is a dense wall of limits, quotas, and billing clauses that an agent must parse mid-task. There is mild redundancy around Groq fallbacks and paid overflow service. Understandable given the tool's complexity, but not tightly structured.

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 25-parameter tool with an output schema and nested host_analysis object, the description covers the operational envelope: size/page/ZIP limits, non-execution guarantees, OCR consent flow, resumability, owner review, and the host-AI handover path. Return-value detail is correctly left to the output schema.

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% across 25 parameters, so the baseline is 3. The description adds relational meaning the schema alone does not: reuse the same _session_id, continue with offset plus expected_source_sha256, collect every range and review before synthesizing, and require_semantic_analysis:true retains an error until analysis is submitted. These pairing rules are genuinely useful beyond per-field docs.

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

Purpose4/5

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

The opening sentence names a specific verb and dual resource: analyze readable UTF-8 text, or manage a private resumable file-analysis job, and later enumerates the job_action verbs. An agent can tell this is a text/file analysis tool distinct from the surrounding continuity, pass, and audit tools. It never names a sibling (e.g. pdf_extract) for boundary-setting, so it stops short of a 5.

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?

Guidance is explicit for internal mode selection: 'Omit to use bounded text/URL analysis', 'use job_action=create...', 'Use analysis_mode=host_ai to skip Zambo inference entirely', and 'When unavailable, YOU can analyze... and submit host_analysis'. It also states prerequisites (active Zambo capability key, Day Pass/Zambo Pass for continuation). It gives no cross-tool comparison against alternatives like pdf_extract, so it is clear context without exclusions.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources