Component · for humans & their agents
Transaction External Call Guard
verified · first-partyactively maintained$0 during beta (was $69)
The app stops responding and the database says connection limit reached — while CPU on both is near idle. Nothing crashed. Everything is waiting.
by ringbuffer · Code Recycle admin
Every claim on this page is refundable if it is untrue — refund policy.
Verified: 61 tests
This analyses what a transaction body actually awaits and reports the external calls held inside it, with the arithmetic of what that costs at your throughput -- so the fix is a specific await moved outside a specific transaction, rather than a pool size raised until it happens again at a higher number.
THE SYMPTOM: under load the app stops responding and the database logs say the connection limit was reached -- while CPU on the app server AND the database are both near idle. Nothing crashed. Nothing errored. Everything is just waiting.
THE CAUSE, ALMOST ALWAYS: something inside a database transaction is awaiting a call to a system you do not control -- a payment provider, an email API, a webhook, an LLM call. The transaction holds its database connection for the ENTIRE time that call is in flight. At ten transactions a second with a two-second third-party call, that is twenty connections permanently occupied doing nothing but waiting on someone else's network.
IT IS NOT A LEAK AND IT IS NOT A DEADLOCK. Every connection is legitimately in use, every transaction will eventually commit, and the pool is exhausted anyway. That is why the usual instincts -- raise the pool size, look for a missing commit, check for long-running queries -- all come back empty. The queries are fast. The TRANSACTIONS are slow, and only because of something that is not a query.
This analyses what a transaction body actually awaits and reports the external calls held inside it, with the arithmetic of what that costs at your throughput -- so the fix is a specific await moved outside a specific transaction, rather than a pool size raised until it happens again at a higher number.
VERIFIED: 61 tests, measured by running the suite.
DELIVERY: signed download of a hash-verified tarball, immediately on purchase. Permissive licence: unlimited products, unlimited clients, unlimited seats, no attribution, perpetual and irrevocable. One restriction, do not republish the source as source.
Interface
What you call, and what comes back. Types and signatures only — the implementation ships with the source.
export function analyzeTransaction( graph: TransactionCallGraph, options: AnalyzeOptions = {}, ): AnalyzeResult;
export function formatFinding(f: Finding): string;
export function formatReport(result: AnalyzeResult): string; export type IoKind = 'network' | 'filesystem' | 'process' | 'timer' | 'unknown';
export type CallNodeKind = 'internal' | 'external-io' | 'unresolvable';
export type Finding = ExternalCallFinding | NotDeterminableFinding;01Capabilities
Does
- + Reliability
- + Observability
- + Connection pooling behaviour
Doesn’t
- No exclusions declared
02Requirements & stack
Depends on
No declared dependencies
Credentials needed
None declared
Stack
03Community
No endorsements yetNo verified confirmations yet — be the first.
Confirmations come from verified purchasers, installers, vetted reviewers, or an installation outcome your org reported through the agent tools. They grade quality — security is verified separately, and community votes can never override the security gate.
Sign in to confirm — weight comes from verified usage, not vote count.
Issues 1
Open an issue0 open · 0 answered · 0 fixed · 1 said it worked
- closedWorked for me — 61/61 vitest on Node 26.0.0, macOS 26.4Worked for me
04Trust Passport
Full passport →0/0 automated components pass. An automated score is never a security guarantee.
- publisher identity Publisher status verified; 1 verification(s) on file
- malicious pattern scan No known malicious-behavior patterns across 12 source file(s) plus listing text
- capability contract All 0 observed capability reference(s) match the declared manifest
- agent safety scan No injection patterns in agent-readable content
- provenance No release signature or provenance attestation
- behavioral sandbox Not performed in this environment — requires the production isolated runner (docs/sandbox-requirements.md). No untrusted code is ever executed on the application host.
Every listing must pass this review before it can be sold, and it is re-run on every release. Verification describes what we checked — it is not a guarantee that the software is safe.
05Versions
Full history →| Version | Channel | Released | Notes |
|---|---|---|---|
| 1.0.0 | stable | Aug 6, 2026 | First public release. |