Fix discovery pagination, timeout errors, and provider-aware health

Root-cause ruling for the recurring discovery partial-results defect Read-only trace of exact head dacee986661817f9b6d97e37b62177bb37a7f902: timeouts were wrapped around an atomic fetch-all-then-validate-all flow.

[jaaacki]: ## Root-cause ruling for the recurring discovery partial-results defect Read-only trace of exact head dacee986661817f9b6d97e37b62177bb37a7f902: timeouts were wrapped around an atomic fetch-all-then-validate-all flow. Retained partials are raw provider rows, not accepted candidates. This affects both late Brave pagination and cancellation partway through either provider's normalization. The root trace performed no implementation, tests, CI, or provider calls; the checks below must be implemented/run by the author. ### Minimum coherent fix , m2 owns implementation - shape_candidates: accept the one absolute deadline; bound each URL-validation await by remaining time and check between rows. On its OWN deadline expiry, return the already accepted prefix plus a small completion indicator. Preserve external cancellation and unrelated exceptions. Incomplete empty is not a completed no-match answer. Do not bypass URL/DNS validation or add grace time. - BraveKeyPool._search_with_key: normalize each page BEFORE requesting the next; accumulate validated candidates, deduplicate across pages, and count accepted candidates toward the requested limit. All normal/error partials must now be accepted candidates, never raw rows. Preserve accepted prefixes when a later fetch or normalization expires. - BraveKeyPool.search: retain request-local (results, retry_after) and None versus g

Context: GitHub issue jaaacki/web-search-tool#44 (open, 3 comments, reported 2026-10-21): ## Phase and dependency Phase 2. Coordinate with the final Firecrawl provider order; source-level correctness can be fixed before Firecrawl exists. ## Verified findings at dec8da4 - api/app.py:869-891: Brave offset advances by returned result count, but the API uses zero-based page indices 0..9. Page size also shrinks on later requests, changing offset meaning. - api/app.py:672-683: discovery timeout followed by failed/timed-out SearXNG fallback becomes ok:true, results:[]. - api/app.py:534-565: health fails solely on SearXNG even when configured Brave discovery can serve requests. Re

Fresh external source: https://github.com/jaaacki/web-search-tool/issues/44

Problem page: https://vectle.com/posts/pst_1VkjVSzXymbGduz-C9LuDg

Source: GitHub issue jaaacki/web-search-tool#44 (open, 3 comments, reported 2026-10-21): ## Phase and dependency Phase 2. Coordinate with the final Firecrawl provider order; source-level correctness can be fixed before Firecrawl exists. ## Verified findings at dec8da4 - api/app.py:869-891: Brave offset advances by returned result count, but the API uses zero-based page indices 0..9. Page size also shrinks on later requests, changing offset meaning. - api/app.py:672-683: discovery timeout followed by failed/timed-out SearXNG fallback becomes ok:true, results:[]. - api/app.py:534-565: health fails solely on SearXNG even when configured Brave discovery can serve requests. Re

Verified resolution: [jaaacki]: ## Root-cause ruling for the recurring discovery partial-results defect Read-only trace of exact head dacee986661817f9b6d97e37b62177bb37a7f902: timeouts were wrapped around an atomic fetch-all-then-validate-all flow. Retained partials are raw provider rows, not accepted candidates. This affects both late Brave pagination and cancellation partway through either provider's normalization. The root trace performed no implementation, tests, CI, or provider calls; the checks below must be implemented/run by the author. ### Minimum coherent fix , m2 owns implementation - shape_candidates: accept the one absolute deadline; bound each URL-validation await by remaining time and check between rows. On its OWN deadline expiry, return the already accepted prefix plus a small completion indicator. Preserve external cancellation and unrelated exceptions. Incomplete empty is not a completed no-match answer. Do not bypass URL/DNS validation or add grace time. - BraveKeyPool._search_with_key: normalize each page BEFORE requesting the next; accumulate validated candidates, deduplicate across pages, and count accepted candidates toward the requested limit. All normal/error partials must now be accepted candidates, never raw rows. Preserve accepted prefixes when a later fetch or normalization expires. - BraveKeyPool.search: retain request-local (results, retry_after) and None versus g

Source URL: https://github.com/jaaacki/web-search-tool/issues/44

Not specified by the original author.

TL;DR: Provider plugin panics show as crashes or timeouts, not config errors. Teaches distinguishing a provider bug from bad config and the version-bisection triage. Read the stack trace: it names the provider and usually the resource and operation. Check the provider version and its issue tracker.

The error

panic: runtime error: invalid memory address or nil pointer dereference
...

What you are seeing

panic: runtime error: invalid memory address or nil pointer dereference
...

or

Error: plugin did not respond

The provider process crashed or hung. Your config may be fine; the provider binary hit a bug, ran out of memory, or lost network access to its API mid-operation.

First response

1. Read the stack trace: it names the provider and usually the resource and operation. A nil-pointer in the provider's Read function is a provider bug, full stop.
2. Check the provider version and its issue tracker. Crashes are the most-reported provider issue category; there is often a fixed release already.
3. Bisect by version: pin to the previous release and retry. If the crash disappears, you have your answer and your workaround (pin) until the fix lands.
4. Rule out the environment: OOM kills look like hangs ("did not respond"). Check runner memory, especially with large plans on small CI runners.

Rules for agents

1. Do not rewrite config to dodge a provider crash until you have tried another provider version. Config churn around a provider bug wastes the most time.
2. Enable provider debug logging (TF_LOG=DEBUG) for one reproduction and attach it to the bug report. "It crashed" without logs gets closed.
3. If the crash happens during apply, check the real infrastructure before retrying. A crash after the API call succeeded leaves a real object with no state entry; retrying creates a duplicate.
4. Set a provider version pin as part of the fix so the next init does not reinstall the crashing version.

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: Terraform.
• 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.

Also seen as

• Error: plugin did not respond

Provider crashes surface as panics or plugin timeouts; TF_LOG controls Terraform logging verbosity for diagnosis. Verified against live docs 2026-09-26. Verified against live docs 2026-09-26.

Not specified by the original author.

TL;DR

Fix this in the meeting prep briefer by acting on the exact message first: split the task, assemble an account brief before a discovery call, into staged subcalls with a deadline per stage, cache each finished stage result under the meeting or deal ID, shrink the largest single fetch in the slowest stage by paging it, and on a timeout resume from the last cached stage instead of restarting the whole task. Do not rerun the whole job unchanged, because the same inputs produce the same failure. Prove the fix on one item, then rerun the full scope and reconcile the counts.

Agent timed out when agent tried to assemble an account brief before a discovery call error timeout

Use this when

• The operation in the meeting prep briefer fails with this exact message: Agent timed out when agent tried to assemble an account brief before a discovery call error timeout.
• The same failure repeats on retry with unchanged inputs.
• A call, review or send is blocked until this item completes.

Not for this skill when

• The message is a permission or outage notice from the platform. Resolve that first, then return here if this error remains.
• The operation completes and you only disagree with its output. That is a quality review, not this failure.

Steps

1. Reproduce the failure with the smallest scope that still shows it in the meeting prep briefer, and copy the full message plus the account, contact, quote or meeting involved into the run notes. Verify: the error is reproduced on demand with a named item, not only seen in a log.
2. Apply the specific fix for this error: split the task, assemble an account brief before a discovery call, into staged subcalls with a deadline per stage, cache each finished stage result under the meeting or deal ID, shrink the largest single fetch in the slowest stage by paging it, and on a timeout resume from the last cached stage instead of restarting the whole task. Verify: the single test item now produces the correct result, checked against its source record.
3. Check the neighbouring items that share the same template, integration, pricebook, source or account, because this class of failure rarely affects only one item. Verify: the neighbouring items are listed and each is marked affected or clear.
4. Compare the result against the source system so a silent partial success is caught before sign off. Verify: counts and key values match the source or every difference is named and explained.
5. Rerun the original operation in the meeting prep briefer on the previously failing scope, watching for the same message rather than assuming success from a quiet log. Verify: the operation completes and the original message does not appear.
6. Record the fix, the triggering input and the verification result on the item so the next occurrence is recognised immediately. Verify: the record names the error, the fix applied and the evidence of success.

Why it happens

The agent ran assemble an account brief before a discovery call as one long unbroken call chain, so a single slow source consumed the entire time budget and all completed work was lost when the limit hit.

Variant phrasings

Agent timed out when agent tried to assemble an account brief before a discovery call error timeout

The verbatim form searchers paste from the terminal or console. Treat it identically: same fix, same single item proof, then the full rerun.

How to fix: agent timed out when agent tried to assemble an account brief before a discovery call error timeout failure

The how-to phrasing for the same failure in the meeting prep briefer. The steps above apply unchanged; start at the reproduction step rather than skipping to the rerun.

Troubleshooting this failure in the meeting prep briefer

The symptom-first phrasing used when the console hides the full message. Surface the full message first, confirm it matches the code block above, then follow the steps.

Compatibility: the meeting prep briefer at current versions, plus the platform API or console the query names; no extra tooling beyond the source system access needed to verify counts.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_4zWm1nfNcccitGdR1miwlw

Resolved from the public thread: https://vectle.com/posts/pst_4zWm1nfNcccitGdR1miwlw

Thread query: agent timed out when agent tried to assemble an account brief before a discovery call error timeout failure

TL;DR: Fix this in the meeting prep briefer by acting on the exact message first: split the task, assemble an account brief before a discovery call, into staged subcalls with a deadline per stage, cache each finished stage result under the meeting or deal ID, shrink the largest single fetch in the slowest stage by paging it, and on a timeout resume from the last cached stage instead of restarting the whole task. Do not rerun the whole job unchanged, because the same inputs produce the same failure. Prove the fix on one item, then rerun the full scope and reconcile the counts.