Coding agent documentation index Fetch the complete documentation index at: https://docs.worklittle.com/docs-agent-manifest.json Use this file to discover all available pages before exploring further.

Retries and spend

Rate limits, retry policy, 402 handling, and controls that keep an agent loop inside budget.

Rate limits

| Surface | Limit |
| --- | --- |
| API key, all authenticated routes | **60 requests per minute** per key, one-minute window |
| Public board profile, list, detail | 120 per minute per IP |
| Public board apply and parse-resume | 20 per minute per IP |
| Public board sitemap | 30 per minute per IP |
| Board GETs with a Bearer key | That key's per-minute limit (60 by default) instead of the per-IP limit |

Throttled requests return 429, error.code RATE_LIMITED, and Retry-After: 60. MCP forwards the same limits as JSON-RPC errors. Reference: Rate limits.

Backoff that respects billing

Retryable:      429, 500, 502, 503, 504, network timeouts
Not retryable:  400, 401, 402, 403, 404
Base delay:     honor Retry-After when present, otherwise 1s
Growth:         2x with full jitter
Max attempts:   3 for metered reads, 5 for free reads
Writes:         retry only with an idempotency key at your layer

Never retry a 402. The org wallet is blocked by a failed charge or a monthly limit. Never blind-retry a side effect. A timeout on POST /jobs/:id/apply may mean the application went through. Check GET /jobs/applied or dedupe on your own request key. For Apply with AI starts, send Idempotency-Key.

Balance and 402

The public API is pay as you go against your organization's wallet. Metered routes return 402 PAYMENT_REQUIRED when a charge fails or a monthly limit is hit. Free routes keep working.

Add a card on Billing and watch Usage. Alert on your spend rate. A 402 means metered features are already down. On 402, disable metered features and keep free ones: GET /jobs/stats, applied-job tracking, public board pages, most ATS reads. Prices: Jobs pricing, Business pricing.

Signed-in AI on worklittle.com is not your API bill

Consumer credit on worklittle.com (Personal monthly free AI credit, Worklittle Instant at $9.99/month with $20/month AI credit) does not govern your integration. API usage is the organization wallet (pay as you go for AI tokens) plus the free quotas and the 60 per minute key limit. Docs chat on docs.worklittle.com does not use your API key.

Budgeting an agent loop

| Control | Implementation |
| --- | --- |
| Cap page size | Clamp `limit` to 10 or 20 in your tool layer |
| Cap searches per turn | Count tool calls. Refuse past the cap and tell the model to use existing results |
| Cap detail fetches | At most three roles per turn |
| Cache within a conversation | Same filters in the same turn return the cached page |
| Prefer free endpoints | `GET /jobs/stats` for counts, tracking routes for pipeline reads |
| Separate keys per surface | Attribute spend per surface |

Enforce these in code, not in the prompt.