Skip to main content
Glama
Sign In

OrgGen AI

by ai.orggen
OverviewDiscussionsAdmin

propose_captable_import

Imports a cap table from a CSV of holders (a spreadsheet saved as CSV), WRITING NOTHING to the schema — files a reviewable proposal (change_proposals, same pattern as propose_changes) that a human then accepts or refuses, row by row, on the review screen. This is the repository's doctrine for any financial data: no direct write. The column mapping (name/shares/type/round…) is guessed automatically from the headers; mapping corrects it. The response carries the review link, meant for the user: nothing is applied without their action on that link.

Input Schema

NameRequiredDescriptionDefault
csvYesRaw CSV content (a spreadsheet's "Save as CSV" export — a single sheet, not a multi-sheet workbook).
titleNoTitle of the proposal (default: "Import cap table (CSV)").
mappingNoColumn → role mapping, to correct the automatic guess — one entry per column to correct (`column` = 0-based index). Omitted = guessed from the headers, like the web screen.
idempotencyKeyYesREQUIRED. A stable key derived from the file (e.g. its name + its date). Replaying with the same key returns the proposal ALREADY created instead of creating a second one.
expectedProjectIdNoUUID of the project targeted by THIS call. ACTUALLY targets this project instead of the session project. Refused without changing anything if access is not granted.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed5 schema fields changed
    • changedInput schema / properties / csv / description
      Previous value: -"Contenu CSV brut (export « Enregistrer sous CSV » d'un tableur — une seule feuille, pas un classeur multi-feuilles)."New value: +"Raw CSV content (a spreadsheet's \"Save as CSV\" export — a single sheet, not a multi-sheet workbook)."
    • changedInput schema / properties / expectedProjectId / description
      Previous value: -"UUID du projet visé par CET appel. Cible RÉELLEMENT ce projet au lieu du projet de session. Refusé sans rien modifier si l'accès n'est pas accordé."New value: +"UUID of the project targeted by THIS call. ACTUALLY targets this project instead of the session project. Refused without changing anything if access is not granted."
    • changedInput schema / properties / idempotencyKey / description
      Previous value: -"OBLIGATOIRE. Une clé stable dérivée du fichier (ex: son nom + sa date). Rejouer avec la même clé rend la proposition DÉJÀ créée au lieu d'en créer une seconde."New value: +"REQUIRED. A stable key derived from the file (e.g. its name + its date). Replaying with the same key returns the proposal ALREADY created instead of creating a second one."
    • changedInput schema / properties / mapping / description
      Previous value: -"Mapping colonne → rôle, pour corriger le devinée automatique — une entrée par colonne à corriger (`column` = index 0-based). Omis = deviné depuis les en-têtes, comme l'écran web."New value: +"Column → role mapping, to correct the automatic guess — one entry per column to correct (`column` = 0-based index). Omitted = guessed from the headers, like the web screen."
    • changedInput schema / properties / title / description
      Previous value: -"Titre de la proposition (défaut : « Import cap table (CSV) »)."New value: +"Title of the proposal (default: \"Import cap table (CSV)\")."
  2. First observed

TDQS

A4.3/5.0
Behavior5/5

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

Annotations only declare readOnlyHint=false and destructiveHint=false; the description goes well beyond that by explaining that a reviewable proposal record is created, that a human accepts or refuses row by row on the review screen, and that nothing is applied until the user acts on the returned link. It also frames this as the repository's doctrine for financial data, which is exactly the kind of 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.

Conciseness4/5

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

Front-loaded with the action and the no-write constraint, and each sentence carries distinct information (proposal model, mapping behavior, returned link). Slightly verbose with emphatic capitalization and a parenthetical aside, but nothing is truly wasted.

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 5-parameter mutation-adjacent tool with no output schema, the description covers the essential mechanics: what is created, that review is required, and that the response carries the review link. It does not touch permissions or failure modes (e.g., what happens with a malformed CSV), leaving a small gap.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents every parameter, including the auto-guess/override behavior of `mapping` and the replay semantics of `idempotencyKey`. The description's mentions of mapping and column roles largely restate that, adding only light framing ('corrects it'), so the baseline 3 is appropriate.

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 (imports a cap table from a CSV of holders) and immediately disambiguates the name: despite 'imports', it writes nothing to the schema and instead files a change proposal. It also names the sibling pattern (propose_changes / change_proposals) so an agent can place it without opening another schema.

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 a clear use context (bulk cap-table import from a spreadsheet export) and explains the proposal-then-review flow that governs when the effect actually lands. It does not mention when not to use it or route to adjacent tools such as validate_captable, so it stops short of explicit alternatives/exclusions.

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

All OrgGen AI toolsInspect
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