Skip to main content
Glama

list_generation_events

Read-onlyIdempotent

Workspace Analytics & Billing: every RUN of ONE Enterprise workspace, one row each, oldest first. Needs a workspace reporting key (wdr_report_…, made by the workspace Owner) for that workspace, or your own sign-in as its Owner or Admin (every run) or Member (only your own runs). A run = one generation job (a retry is a new run), or one step charged outside a job (a Cinematic Ads analysis, AI step or export, a Drama Studio chat turn, a recap, a Wan Motion Control step, subtitles, a Frame Studio export…), counted once, so the runs' credits add up to the Usage page. Each row of data: attempt_id (stable: the job id, or the step's ledger ref), job_id (null for a step outside jobs), source_channel (web, api_key, oauth, mcp; api = API key or OAuth before 10/10/2026; null = not recorded), actor_type (always human: every run is made by a person), user_id, team (the person's team when the run finished; "unclassified" = none), project_id and app_name (the Cinematic Ads project's app name), service: feature (a step outside jobs: what it was charged for, e.g. drama_chat), model_id (null for a step outside jobs) and model_name as the web shows them, provider (always "wandura"); output: kind (image / video / audio), count (files made; 0 unless succeeded), video_duration_seconds (measured from the file, not rounded) and duration_bucket, requested_duration_seconds (the length asked for) and requested_duration_bucket — buckets 0-15, 15-30, 30-60, 60-120, 120+ seconds, each upper bound included (15 is 0-15, 15.07 is 15-30), unknown = not measured or not asked; resolution (shorter side, e.g. 720p) and aspect_ratio (e.g. 9:16), both measured; preview_url (the CDN link: anyone with it can open it); status (succeeded, failed, running; there is no cancelled: a run the provider cancelled is failed), error_code, retry_of (the run this one retried); request_submitted_at, generation_started_at (+ generation_started_quality: exact, or approximate for models that do not report a start, which can be off by up to 2 minutes), generation_finished_at (when the result was ready; a run that failed at submit: when it was refunded); queue_time_seconds = started − submitted, generation_time_seconds = finished − started, request_to_result_seconds = finished − submitted, per run (never add parallel runs together); credits: gross_debit, refunded, net (= gross_debit − refunded), reserved (still held by a run in progress); cost_effective_usd = net × meta.cost_basis.effective_cents_per_credit (the workspace's one effective price), cost_list_usd = net × the list price (1 credit = $0.01); updated_at. A missing time is null, never 0, and every wait computed from it is null; meta.missing_fields lists each such field with the number of runs and why (one entry per cause). Range: a run belongs to the day it finished (a run still in progress: the day it was submitted). from (first day, included) and to (the day after the last, excluded), YYYY-MM-DD in Asia/Ho_Chi_Minh, at most 366 days; default the last 30 days including today. Paging: up to limit rows (1–1,000, default 1,000); when meta.next_cursor is not null pass it as cursor (with the same range and filters) for the next page. Sync: updated_since = the latest updated_at you stored returns only runs changed since. Free; at most 60 calls a minute per key or person. CSV: GET /api/v1/workspaces/{workspace_id}/reports/list_generation_events?format=csv (next page cursor in the X-Next-Cursor header).

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": {
      "workspace_id": "<workspace_id>"
    },
    "name": "list_generation_events"
  }
}

Input Schema

NameRequiredDescriptionDefault
toNoThe day after the last, excluded (YYYY-MM-DD, Asia/Ho_Chi_Minh). Default: tomorrow. At most 366 days after `from`.
appNoOnly runs of this app (app_name: the Cinematic Ads project's app name).
fromNoFirst day, included (YYYY-MM-DD, Asia/Ho_Chi_Minh). Default: 30 days before `to`.
teamNoOnly runs of this team, as the report names it ("unclassified" = people with no team).
limitNoRows per page.
cursorNometa.next_cursor of the previous page, unchanged.
statusNoOnly runs in this state.
channelNoOnly runs that came this way.
featureNoOnly runs of this feature (a feature from list_services, or what a step outside jobs was charged for: cinematic_ads_analysis, cinematic_ads_ai, cinematic_ads_export, drama_chat, …).
user_idNoOnly this person's runs (a user id from list_workspace_members).
model_idNoOnly runs of this model (a model_id from list_services; older names of the same model are included).
project_idNoOnly runs of this project (a Cinematic Ads or Drama Studio project id).
resolutionNoOnly videos of this measured resolution, e.g. 720p (the shorter side in pixels).
output_kindNoOnly runs making this kind of output.
workspace_idYesThe workspace: the one the reporting key belongs to, or one you are a member of.
updated_sinceNoOnly runs changed after this time (to the second), e.g. the latest updated_at you stored.
duration_bucketNoOnly videos whose measured length is in this bucket.
requested_duration_bucketNoOnly videos whose requested length is in this bucket.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A4.5/5.0
Behavior5/5

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

Beyond the readOnly/idempotent annotations, it discloses cost (free), throttling (60 calls/minute per key or person), auth paths and scope limits, null-vs-zero time semantics, that preview_url is publicly readable, that 'there is no cancelled' state, that retries are separate runs, and that meta.missing_fields explains gaps. These are exactly the operational facts an agent needs before calling.

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 purpose and auth are front-loaded, but the 'Each row of data' section is a single enormous run-on enumeration that is hard to scan, and paging/limit rules are restated from the schema. Nearly every detail is substantive, but the lack of structure (no breaks by concern) hurts usability.

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?

With 18 parameters, no output schema, and complex billing semantics, the description carries the full burden of documenting the returned row shape, and it does so field by field, including derived cost figures and time-delta definitions. Nothing an agent needs to call or interpret this tool is missing.

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 schema already carries baseline meaning, but the description adds real value: bucket boundary inclusivity (15 is 0-15, 15.07 is 15-30), day attribution by finish time (submit time for running runs), from inclusive / to exclusive, and that cursor must be sent with the same range and filters. Some schema-level details (limit bounds, cursor definition) are simply repeated rather than extended.

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 opening sentence names a specific verb (list), a specific resource (generation runs of one Enterprise workspace), the grain (one row per run), and the ordering (oldest first). It also crisply defines what a 'run' is, including retries and out-of-job steps, so the agent cannot confuse it with get_usage_summary or list_credit_transactions.

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?

It states prerequisites precisely (workspace reporting key from an Owner, or Owner/Admin/Member sign-in with Member limited to own runs), the default range, the 366-day cap, paging via meta.next_cursor, incremental sync via updated_since, and rate limits. What it lacks is explicit routing language naming a sibling (e.g. 'use get_usage_summary for totals'), leaving the Usage-page reconciliation only implied.

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