Repair Content-Encoding Bugs

The problem

HTTP body transformations often fail when libraries and wrappers disagree on state ownership. This leads to double-decoding errors, "decompression bombs" bypassing size limits, and stale headers that claim content is encoded when it has already been stripped.

What it does

• Identifies and repairs ordered decoder stack failures for gzip, br, deflate, and zstd.
• Enforces bounded decoding limits to prevent resource exhaustion from high expansion ratios.
• Syncs representation metadata like Content-Length, ETag, and signatures with the actual transformation state.
• Fixes streaming backpressure and cleanup leaks in corrupt or cancelled streams.
• Generates deterministic local fixtures to replicate encoding edge cases without hitting production APIs.

Frameworks & tools

Works with any language or framework using standard HTTP encoding layers. Specifically addresses implementations involving gzip, brotli (br), deflate, and zstd. Uses Read, Edit, and Bash tools for repository-level patching.

Why this beats prompting it yourself

Manually fixing encoding bugs often results in "whack-a-mole" where fixing a decoder breaks the Content-Length or ETag. This skill follows a strict resource register and synchronization workflow to ensure all metadata layers are updated atomically and safely.

Use cases

• Fixing double-decompression bugs where both a client and a middleware try to unzip a response.
• Patching security vulnerabilities where compressed input exceeds memory limits.
• Correcting order-of-operation errors in stacked HTTP encodings.
• Updating stale ETag and Content-Length headers after body transformation.

Known limitations

Does not modify content negotiation logic or public security limits. Cannot access live production environments, secrets, or remote caches.