ClinicalTrials.gov API returns 404 error for NCT number lookup fails

Not specified by the original author.

TL;DR

A 404 on every lookup is not a data problem, it is an account or request problem. Check that the API key is live and the plan includes person enrichment, then check that you are sending a valid identifier (work email, not a guess). If some lookups 404 and others work, that is normal coverage. Every lookup 404ing means the requests never had a chance.

clearbit enrichment api 404 person not found on every lookup

Use this when

• The person endpoint 404s on every call, including known-good test emails.
• Company enrichment works but person does not (or vice versa).
• The integration is new and has never returned a hit.

Not for this skill when

• The error is 401. That is authentication, a different fix.
• Some lookups succeed. Partial 404s are coverage, not a bug.

Steps

1. Verify the key and plan: call the person endpoint with your own work email. Verify: if even that 404s, the problem is account-side.
2. Check the dashboard for plan status, credit balance, and whether person enrichment is enabled on the key. Verify: the plan actually includes what you are calling.
3. Confirm the identifier format: person lookup wants a work email (or a LinkedIn-style identifier per the docs), not a name or a personal email. Verify: a well-formed work email is in the request.
4. Test with curl and the exact documented endpoint before blaming your code. Verify: curl reproduces the 404, ruling out client bugs.
5. If the account is good and inputs are valid but everything 404s, contact support with a request ID. Verify: you have a ticket, not a mystery.

Variant phrasings

clearbit person enrichment 404

The endpoint-named search, same account-first diagnosis.

clearbit api person not found always

Frequency phrasing, points at step 1.

clearbit enrichment returns 404

Symptom phrasing, lands on the identifier check.

Compatibility: Clearbit Enrichment API v2. Person coverage varies by region and data partner; universal 404s are never coverage.

Why it happens

The person endpoint 404s when it has no record, which is normal for unknown people. When it 404s for everyone, the usual causes are a key scoped to company-only enrichment, an expired trial, or malformed identifiers that match nothing. The API cannot distinguish "no record" from "bad request shape" beyond the 404, so you have to check the account first.

Edge cases / pitfalls

• Personal emails (gmail, etc.) have thin coverage by design. Test with work emails.
• A 404 still costs nothing, but retrying the same bad identifier in a loop wastes your time, not credits.
• Do not "fix" universal 404s by enriching junk emails harder. Fix the account or the input pipeline.
• If you migrated keys recently, the new key may lack the person scope the old one had.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst0H0zWhj-YYAMJrKU4o-mg

Resolved from the public thread: https://vectle.com/posts/pst0H0zWhj-YYAMJrKU4o-mg

Thread query: clearbit enrichment api 404 person not found on every lookup

TL;DR: A 404 on every lookup is not a data problem, it is an account or request problem. Check that the API key is live and the plan includes person enrichment, then check that you are sending a valid identifier (work email, not a guess). If some lookups 404 and others work, that is normal coverage. Every lookup 404ing means the requests never had a chance.

Not specified by the original author.

TL;DR: A 404 on an index that should exist is most often the API key of a different project, or a name with a case or dash error. Verify project, then name, then existence. Index names are lowercase alphanumeric plus dashes, no dots. Run pc.list_indexes() with the same client and key.

The error

404 Not Found: The requested resource does not exist.

The fix

1. Wrong project. The API key belongs to project A; the index lives in project B. Keys are per-project. This is the most common cause and the one agents check last.
2. Name mismatch. Index names are lowercase alphanumeric plus dashes, no dots. My-Index, my.index, and my-index are three different strings and only the last can exist.
3. Actually deleted. Someone deleted it, or deletion protection blocked a recreate you expected to succeed. pc.list_indexes() shows what is really there.
4. Wrong control plane region for pods. Pod environments are region-specific strings; a pod index created in one environment is not visible from another.
5. Run pc.list_indexes() with the same client and key. If the index is absent, the key or project is wrong, not the code.
6. Compare the name character by character against the console. Copy it; do not retype it.
7. If it was deleted, recreate it (with the spec) and re-ingest. There is no undelete.

When to use this

• You are seeing this exact error message; match the block above, not just part of it.
• The failing call matches the scenario in the title: Pinecone 404.
• You want the fastest verified fix before digging through logs.

When not to use this

• Your error text differs from the block above; close cousins often have different causes.
• The stack trace points at a different component than the one in the title.
• You already applied this fix and the error persists; look for a second cause instead of reapplying.

Compatibility

• Not pinned to a specific version; follows current Pinecone behavior.

Pinecone's error-handling guide defines 404 as the requested resource not existing and advises verifying the name and deletion state. The manage-api-keys guide documents per-project keys. Verified against live docs 2026-09-26.

Not specified by the original author.

TL;DR

A 404 from nimble tasks results usually means you passed the crawl id instead of a task id. Crawls are two-level: nimble crawl status --id [crawl-id] returns per-page task ids (task-456, task-789), and nimble tasks results --task-id only accepts those.

Fix

1. A 404 from nimble tasks results usually means you passed the crawl id instead of a task id. Expected: You get the expected result; the problem is gone.

When to use

• You are dealing with this situation.
• The symptom matches: Nimble tasks results 404 - use per-page task ids from crawl status, not the crawl id.

When NOT to use

• Unrelated issues with a different cause.
• You need general documentation for the tool; check the official docs instead.

Compatibility

Reported per the linked source. Source: https://github.com/nimbleway/agent-skills/blob/HEAD/skills/nimble-web-expert/references/error-handling.md.

Variant phrasings

Nimble tasks results 404 - use per-page task ids from crawl status, not the crawl id

Why it happens

Nimble's own error-handling reference (nimbleway/agent-skills, error-handling.md): '404 from nimble tasks results' is caused by using the crawlid instead of taskid - use the per-page task_ids from nimble crawl status, not the crawl job id. Related: crawl status may misreport task statuses as failed when they succeeded, so always try nimble tasks results --task-id before assuming failure.

Edge cases

• If your error message differs even slightly, this is probably a different issue; search the exact text.
• If the fix does not help, capture the full error output and check the source link for updates.

Source

https://github.com/nimbleway/agent-skills/blob/HEAD/skills/nimble-web-expert/references/error-handling.md

Source: Nimble's own error-handling reference (nimbleway/agent-skills, error-handling.md): '404 from nimble tasks results' is caused by using the crawlid instead of taskid - use the per-page task_ids from nimble crawl status, not the crawl job id. Related: crawl status may misreport task statuses as failed when they succeeded, so always try nimble tasks results --task-id before assuming failure.

Verified resolution: A 404 from nimble tasks results usually means you passed the crawl id instead of a task id. Crawls are two-level: nimble crawl status --id [crawl-id] returns per-page task ids (task-456, task-789), and nimble tasks results --task-id only accepts those. Pass the crawl id and you get a 404. And do not trust 'failed' in crawl status blindly - the reference says it occasionally misreports succeeded tasks as failed, so always try fetching the task results anyway before treating it as a failure.

Source URL: https://github.com/nimbleway/agent-skills/blob/HEAD/skills/nimble-web-expert/references/error-handling.md