What causes "validation_failed" in rohitg00/agentmemory?

Retry — nondeterministic LLM output may validate on a second attempt.

validation_failed

validation_failed

Error message

validation_failed

What it means

After the provider returns a summary, mem::summarize validates the parsed summary XML against schema/rules (validation.result). If validation reports errors, the function logs them and returns success:false with error 'validation_failed', refusing to persist an invalid summary to KV.summaries.

Source

Thrown at src/functions/summarize.ts:355

          concepts: summary.concepts,
        };
        const validation = validateOutput(
          SummaryOutputSchema,
          summaryForValidation,
          "mem::summarize",
        );

        if (!validation.valid) {
          const latencyMs = Date.now() - startMs;
          if (metricsStore) {
            await metricsStore.record("mem::summarize", latencyMs, false);
          }
          logger.warn("Summary validation failed", {
            sessionId,
            errors: validation.result.errors,
          });
          return { success: false, error: "validation_failed" };
        }

        const qualityScore = scoreSummary(summaryForValidation);

        await kv.set(KV.summaries, sessionId, summary);
        await safeAudit(kv, "compress", "mem::summarize", [sessionId], {
          title: summary.title,
          observationCount: compressed.length,
        });

        const latencyMs = Date.now() - startMs;
        if (metricsStore) {
          await metricsStore.record(
            "mem::summarize",
            latencyMs,
            true,
            qualityScore,
          );
        }

View on GitHub (pinned to e04ba88819)

Solutions

Retry — nondeterministic LLM output may validate on a second attempt.
Use a stronger model that reliably follows the summary XML schema.
Update agentmemory (or the model prompt) so the expected schema matches the produced output.
Check the logged validation.result.errors to see exactly which fields failed and adjust configuration accordingly.

Example fix

// before
await sdk.trigger({ function_id: "mem::summarize", payload: { sessionId } }); // ignored validation_failed
// after
const r = await sdk.trigger({ function_id: "mem::summarize", payload: { sessionId } });
if (r.error === "validation_failed") await sdk.trigger({ function_id: "mem::summarize", payload: { sessionId, model: "stronger-model" } });

Defensive patterns

Strategy: retry

Validate before calling

// confirm the model in config matches the one your schema was validated against
if (!/strong-model/.test(provider.model)) console.warn("model may not produce schema-valid summary XML");

Type guard

function isValidationFailed(res: unknown): res is { success: false; error: "validation_failed" } {
  return typeof res === "object" && res !== null && (res as any).error === "validation_failed";
}

Try / catch

const res = await sdk.trigger({ function_id: "mem::summarize", payload: { sessionId } });
if (isValidationFailed(res)) {
  const retry = await sdk.trigger({ function_id: "mem::summarize", payload: { sessionId } });
  if (isValidationFailed(retry)) escalateOrFallback(retry);
}

Prevention

Use models known to follow strict XML output instructions.
Keep agentmemory and its expected summary schema in sync with your prompts.
Retry automatically on validation_failed before surfacing to users.
Read logged validation errors to target prompt fixes.

When it happens

Trigger: registerSummarizeFunction: the LLM output parses to a summary but validation.result.valid is false (missing required fields like title, malformed structure, constraint violations), so the handler returns error 'validation_failed'.

Common situations: Smaller/weaker model omitting required summary fields; prompt/model changes altering XML shape; schema updated in a new agentmemory version while responses follow the old format; truncation dropping required trailing elements.

Understand the failure class

Background: Schema validation failed / invalid input schema: payload rejected because its shape doesn't match the expected schema — this error's family across 28 libraries.

Related errors

AI-assisted analysis of rohitg00/agentmemory@e04ba88819 (2026-08-30). Data as JSON: /api/errors/5076b5f688e53571. Report an issue: GitHub.