Walkthrough

This PR adds configurable retry and backoff behavior for the LLMWhisperer v2 client by introducing three environment variables, updating the client dependency version, and propagating these configuration parameters through the adapter initialization chain.

Changes

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

Looks like some lock files were not updated in a long time like workflow-execution and workers. That could be the reason for such huge file diffs.

Test Results

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

Greptile Summary

This PR adds configurable retry backoff parameters (max_retries, retry_min_wait, retry_max_wait) to the LLMWhispererClientV2 constructor in sdk1 to gracefully handle transient HTTP errors (429, 5xx), and bumps the llmwhisperer-client minimum version from 2.2.1 to 2.6.2 which exposes these new constructor parameters.

• New env vars ADAPTER_LLMW_MAX_RETRIES, ADAPTER_LLMW_RETRY_MIN_WAIT, and ADAPTER_LLMW_RETRY_MAX_WAIT are read at module import time via WhispererDefaults class-level attributes and forwarded directly to the client constructor — consistent with how WAIT_TIMEOUT and LOGGING_LEVEL are already handled.
• prompt-service/sample.env is correctly updated; however, backend/sample.env (which also contains an ADAPTER_LLMW_* block) was not updated despite the PR description claiming both files were changed.
• The WhispererEnv docstring's Attributes section was not updated to document the three new constants.

Confidence Score: 4/5

• Safe to merge — the logic change is minimal and backward-compatible; the only gap is a missing update to backend/sample.env.
• The core code change (three new constructor kwargs forwarded from env-driven defaults) is correct, well-isolated, and backward-compatible with sensible defaults. The only concrete gap is backend/sample.env not being updated, which is a documentation/discoverability issue rather than a runtime bug.
• backend/sample.env — needs the three new retry env var entries added to match prompt-service/sample.env.

Important Files Changed

Sequence Diagram

sequenceDiagram
    participant Env as Environment Variables
    participant Defaults as WhispererDefaults
    participant Helper as LLMWhispererHelper
    participant Client as LLMWhispererClientV2
    participant API as LLMWhisperer API

    Env->>Defaults: ADAPTER_LLMW_MAX_RETRIES (default: 3)
    Env->>Defaults: ADAPTER_LLMW_RETRY_MIN_WAIT (default: 1.0s)
    Env->>Defaults: ADAPTER_LLMW_RETRY_MAX_WAIT (default: 60.0s)

    Helper->>Client: LLMWhispererClientV2(max_retries, retry_min_wait, retry_max_wait)
    activate Client

    Helper->>Client: client.whisper(...)
    Client->>API: POST /whisper
    API-->>Client: HTTP 429 / 5xx (transient error)
    Note over Client: exponential backoff retry<br/>(up to max_retries times,<br/>wait between min_wait and max_wait)
    Client->>API: POST /whisper (retry)
    API-->>Client: HTTP 200 OK
    Client-->>Helper: response
    deactivate Client

This is a comment left during a code review.
Path: backend/sample.env
Line: 152

Comment:
**Missing retry env vars in backend sample.env**

The PR description states "Updated `sample.env` files for both backend and prompt-service," but `backend/sample.env` was not changed in this PR. It contains the same `ADAPTER_LLMW_*` block as `prompt-service/sample.env` but the three new retry variables are absent:

```
ADAPTER_LLMW_POLL_INTERVAL=30
ADAPTER_LLMW_MAX_POLLS=1000
ADAPTER_LLMW_STATUS_RETRIES=5
# ADAPTER_LLMW_MAX_RETRIES, ADAPTER_LLMW_RETRY_MIN_WAIT, ADAPTER_LLMW_RETRY_MAX_WAIT are missing
```

This means developers deploying or configuring the backend service using `backend/sample.env` as a reference won't discover the new tunable retry parameters. Please add the same three entries to `backend/sample.env` that were added to `prompt-service/sample.env`.

How can I resolve this? If you propose a fix, please make it concise.

---

This is a comment left during a code review.
Path: unstract/sdk1/src/unstract/sdk1/adapters/x2text/llm_whisperer_v2/src/constants.py
Line: 37-44

Comment:
**`WhispererEnv` docstring not updated for new env vars**

The class docstring lists `WAIT_TIMEOUT` and `LOG_LEVEL` under `Attributes`, but the three newly added env variables (`MAX_RETRIES`, `RETRY_MIN_WAIT`, `RETRY_MAX_WAIT`) are not mentioned. This makes the docstring incomplete and inconsistent with the actual class state.

```suggestion
class WhispererEnv:
    """Env variables for LLMWhisperer.

    Can be used to alter behaviour at runtime.

    Attributes:
        WAIT_TIMEOUT: Timeout for the extraction in seconds. Defaults to 300s
        MAX_RETRIES: Max retry attempts for transient HTTP errors. Defaults to 3
        RETRY_MIN_WAIT: Min backoff wait in seconds between retries. Defaults to 1.0
        RETRY_MAX_WAIT: Max backoff wait in seconds between retries. Defaults to 60.0
        LOG_LEVEL: Logging level for the client library. Defaults to INFO
    """
```

How can I resolve this? If you propose a fix, please make it concise.

_{Last reviewed commit: 6df4273}