All Articles
When Thinking Is Always On, Prefill Quietly Stops Working — Fixing Streaming and Token Budgets for Fable 5
Fable 5 thinks by default. Prefill no longer applies, the first streamed block isn't text, and max_tokens has to leave room for reasoning. Here is how I fixed those three broken assumptions in my own automated publishing pipeline.
When the Same Model Has a Different Name Everywhere — Designing a Cross-Provider Model Identity Resolver for Claude
Now that Fable 5 is available on the API, Bedrock, and Vertex at once, the same model carries a different identifier on each. Here is how to untangle hardcoded model strings with a small resolver that maps logical names to physical IDs, carries capability flags, and verifies identifiers at startup.
An Unattended Agent That Wakes Up to a Blank Machine Every Time
A scheduled, unattended agent wakes up on a fresh disposable machine every run: paths it cannot write to, a filesystem that has not finished booting, a working directory that has vanished. Here is how to design state recovery into the first thirty seconds, drawn from real operational logs.
When Claude API Prompt Caching Quietly Stops Hitting in Production — Field Notes on TTL and Measured Savings
Prompt caching works beautifully the day you ship it, then quietly stops hitting in production. The five things that break the prefix, how to choose between 5-minute and 1-hour TTL, and how to measure real savings from usage instead of guessing.
Your Claude Files API Storage Is Quietly Filling Up — Dedup With a Content-Hash Ledger and Reap the Orphans
Use the Files API in an automated pipeline and the same file gets uploaded again and again while orphaned files pile up unnoticed. Here is a content-hash dedup ledger plus an orphan GC design, with working code.
Drop Your Static Claude API Keys: Moving CI and Production to Keyless Auth with Workload Identity Federation
Workload Identity Federation is now generally available on the Claude Platform. This guide walks through replacing long-lived sk-ant- keys with short-lived OIDC tokens, including keyless GitHub Actions auth, the migration steps, and token refresh design.
When Your Claude API Cost Math Doesn't Match the Bill: Accounting for the Four Token Buckets
Turn on prompt caching and your homegrown cost tally drifts from the console bill. Here is how to weight the four token buckets the usage object returns and build a ledger you can reconcile.
Claude API Streaming Breaks the "Everything Arrives" Assumption — Field Notes on Recovering from Partial Failure
Once concurrency climbs, Claude API streams disconnect mid-response, replay events, and emit half-finished tool arguments. Treating partial failure as the norm rather than an anomaly, here is how I rebuilt the implementation and monitoring to recover quietly.
Putting a Ceiling on the pause_turn Loop: Running Long Server Tools Safely Unattended
A production design for continuing pause_turn safely in unattended runs, where long server tools like web_search and code execution are involved. Covers branching all four stop_reason values in one loop, capping continuations and wall-clock time, and accumulating usage across paused segments.
Surviving the 90-Second Code Execution Cell Limit with Checkpointed Chunking
Claude's code execution tool now enforces a 90-second per-cell limit. Here is how to keep a long batch from getting cut off there: persist progress to the container filesystem and resume across cells, with working code for timing, idempotent checkpoints, and knowing when to offload.
Connecting Managed Agents to Services You Don't Want to Expose: MCP Tunnel Design
How to connect Claude Managed Agents to an internal MCP server that is never exposed to the public internet. We cover the MCP tunnel, self-hosted sandboxes, authorization boundaries, and graceful degradation when things break.
Don't Carry Search Results Twice: Trimming Consumed Blocks with response_inclusion
When an agent runs dynamic filtering, output tokens balloon because the raw search-result blocks a code execution call already consumed get echoed back into the response. Here is when response_inclusion: excluded is safe to use, when you must keep full, with implementation and a decision table.