tinyhumansai/openhuman · error

learning_enrich_profile: {e:#}

Error message

learning_enrich_profile: {e:#}

What it means

Thrown by learning_enrich_profile when run_linkedin_enrichment() fails — a multi-stage external pipeline: it needs a signed-in integration client ('no integration client — user not signed in'), goes through Composio/Apify (LinkedIn scrape, 'Apify run failed' / empty items) and Gmail evidence fetch, then writes PROFILE.md and persists the profile into the memory store. Any stage failure bubbles up here with the full {e:#} chain naming the broken stage.

Source

Thrown at src/openhuman/agent/learning/tools.rs:603

    fn external_effect(&self) -> bool {
        true
    }

    async fn execute(&self, args: serde_json::Value) -> anyhow::Result<ToolResult> {
        log::debug!("[tool][learning] enrich_profile invoked");
        let preset = args
            .get("profile_url")
            .and_then(serde_json::Value::as_str)
            .map(str::to_string);
        let config = config_rpc::load_config_with_timeout()
            .await
            .map_err(|e| anyhow::anyhow!("learning_enrich_profile: {e}"))?;
        let result =
            crate::openhuman::agent::learning::linkedin_enrichment::run_linkedin_enrichment(
                &config, preset,
            )
            .await
            .map_err(|e| anyhow::anyhow!("learning_enrich_profile: {e:#}"))?;
        Ok(ToolResult::success(serde_json::to_string(&json!({
            "profile_url": result.profile_url,
            "profile_data": result.profile_data,
            "stages": result.stages,
            "log": result.log,
        }))?))
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::openhuman::tools::traits::ToolScope;

    #[test]
    fn names_and_levels() {
        assert_eq!(LearningListFacetsTool.name(), "learning_list_facets");
        assert_eq!(

View on GitHub (pinned to a221052e0d)

Solutions

1. Read the {e:#} chain first — it names the exact stage ('no integration client', 'Apify run failed', 'GMAIL_FETCH_EMAILS failed', store/write errors)
2. Sign in / reconnect the LinkedIn integration in Connections, then retry
3. Verify Composio/Apify credentials and network reachability
4. For a profile that yields empty items, pass an explicit profile_url and retry, or fall back to learning_save_profile with pasted markdown

Example fix

// before
tool: learning_enrich_profile {} // -> learning_enrich_profile: no integration client — user not signed in
// after: connect LinkedIn in Connections UI, then
tool: learning_enrich_profile {"profile_url":"https://www.linkedin.com/in/<handle>"}

// Preconditions from the pipeline's own failure modes:
// 1) integration client must exist (user signed in) — check Connections status for LinkedIn
// 2) Composio/Apify credentials configured
// 3) workspace writable (learning_save_profile with empty-ish markdown succeeds)
// Only then invoke learning_enrich_profile.

match tool_exec("learning_enrich_profile", args).await {
    Err(e) => {
        let msg = e.to_string();
        if msg.contains("no integration client") || msg.contains("not signed in") {
            return Err(e); // auth — retrying won't help until the user connects
        }
        if msg.contains("Apify") || msg.contains("GMAIL") || msg.contains("timeout") {
            tokio::time::sleep(Duration::from_secs(5)).await;
            return tool_exec("learning_enrich_profile", args).await; // transient upstream
        }
        Err(e)
    }
    ok => ok,
}

When it happens

Trigger: User not signed in / LinkedIn integration not connected; missing or invalid Composio/Apify credentials; Apify run failing or returning zero items for a private/hidden profile; network outage mid-scrape; memory client unavailable at persist time; PROFILE.md write failure.

Common situations: First enrichment attempt before connecting the LinkedIn integration; expired third-party tokens; scraping profiles that require login walls; flaky upstream Apify jobs.

Related errors

• learning_save_profile: summarisation failed: {e:#}
• integrationId must be a 24-char hex id
• integration id is required
• tick failed: {e}
• learning_update_facet: {e:#}

AI-assisted analysis of tinyhumansai/openhuman@a221052e0d (2026-08-16). Data as JSON: /api/errors/ed5a590578c72f15. Report an issue: GitHub.