Question
When users run lark-cli, it is not immediately clear which commands consume Feishu/Lark OpenAPI quota and where a personal user or app developer can inspect the remaining usage. This is especially important for AI-agent workflows, where a single user action may trigger multiple CLI commands.
Could the project clarify the quota behavior in README/docs/help?
Specific points to clarify
- Do executed shortcut commands, generated API commands, and raw
lark-cli api calls count against the tenant/app OpenAPI monthly call quota in the same way as SDK/curl requests?
- Are local-only commands such as
--help, schema, config, profile, and --dry-run excluded from quota because they do not execute OpenAPI requests?
- For paginated commands such as
--page-all, should users expect quota usage to equal the actual number of OpenAPI requests/pages fetched?
- Where can an individual user or app developer view quota usage? Current understanding is that enterprise admins can view it in Feishu Admin Console -> Billing/费用中心 -> Entitlements/权益数据, but ordinary personal users may not have access.
- Should quota/rate-limit errors such as
99991403 or 99991400 get a dedicated README/FAQ entry or CLI hint?
Why this matters
lark-cli is often used by AI agents. Users may not realize that agent-driven read/write workflows can consume OpenAPI call quota, especially when commands auto-paginate or when shortcuts perform multiple API calls internally. A short official explanation would help users choose --dry-run, pagination limits, and retry behavior more safely.
Current understanding
--dry-run prints the planned request without executing it, so it should not consume quota.- Real OpenAPI requests issued by
lark-cli likely count the same as requests made through the SDK or direct HTTP calls.
- Quota visibility appears to be tenant/admin-oriented rather than personal-user-oriented.
Please correct the above if any part is inaccurate.
Question
When users run
lark-cli, it is not immediately clear which commands consume Feishu/Lark OpenAPI quota and where a personal user or app developer can inspect the remaining usage. This is especially important for AI-agent workflows, where a single user action may trigger multiple CLI commands.Could the project clarify the quota behavior in README/docs/help?
Specific points to clarify
lark-cli apicalls count against the tenant/app OpenAPI monthly call quota in the same way as SDK/curl requests?--help,schema,config,profile, and--dry-runexcluded from quota because they do not execute OpenAPI requests?--page-all, should users expect quota usage to equal the actual number of OpenAPI requests/pages fetched?99991403or99991400get a dedicated README/FAQ entry or CLI hint?Why this matters
lark-cliis often used by AI agents. Users may not realize that agent-driven read/write workflows can consume OpenAPI call quota, especially when commands auto-paginate or when shortcuts perform multiple API calls internally. A short official explanation would help users choose--dry-run, pagination limits, and retry behavior more safely.Current understanding
--dry-runprints the planned request without executing it, so it should not consume quota.lark-clilikely count the same as requests made through the SDK or direct HTTP calls.Please correct the above if any part is inaccurate.