Agent Engineering Wiki

The substrate churns across several layers, and each is a drift source:

• Frameworks ship fast and regress: LangGraph 1.2.6 had to fix nested subgraphs inheriting the parent checkpoint namespace — a regression introduced two releases earlier in 1.2.3 — meaning anyone who upgraded into that window silently got broken checkpointing without touching their own code.
• Agent SDKs move almost daily: the Claude Agent SDK for Python ships releases whose entire changelog is "updated the bundled Claude CLI," so the executable your agent runs on changes underneath a patch-level dependency bump. That cadence has not let up: in a single recent week the SDK rolled from 0.2.105 through 0.2.110, every release doing nothing but advancing the bundled CLI (2.1.183 → 2.1.191), while the CLI itself shipped its own stream of patch releases (claude-code 2.1.190, "bug fixes and reliability improvements") — so a team that pinned only the SDK version still saw the binary under it move roughly daily.
• Models get deprecated out from under running agents — Claude Code now emits a warning when the requested model is deprecated, making model-upgrade drift an explicit, surfaced signal rather than a silent behavior change — and the same release hardened auto-mode safety (blocking destructive git commands), a reminder that the harness's *defaults* drift too.
• Serving runtimes drift in performance and output: vLLM v0.23.0 is another "hardening and optimization pass" on DeepSeek-V4 across backends, the kind of change that can move latency, throughput, and sampling behavior without a model swap, and the drift can be outright breaking, not just behavioral — Triton Inference Server's 2.70.0 release drops Windows support entirely and changes how its Python client handles BF16 (now requiring ml_dtypes), so a runtime bump can remove a deployment target or break client code that never touched the model.

The field is starting to give operators levers — LangGraph's CLI now supports declaring *compatible API version ranges* — but the default posture is still "track latest," which is exactly how drift gets in.

Fresh evidence that the bundled-CLI drift is structural, not a one-off: across a single week the Claude Agent SDK for Python went 0.2.105 → 0.2.110 with every release just bumping the vendored CLI (2.1.183 → 2.1.191), alongside the CLI's own patch stream (claude-code 2.1.190) — concrete proof that pinning a direct dependency leaves the executable underneath moving almost daily.

A new mitigation targets vulnerability drift specifically: deptrust checks an agent's resolved dependency versions across a dozen package ecosystems against known-vulnerability databases, catching the case where drift moves a project into an unsafe version rather than just a behaviorally different one.

The countervailing signal still holds: drift is being made visible at the seams (Claude Code warns on deprecated models, LangGraph's CLI declares compatible API version ranges), early steps toward treating the substrate as a versioned contract rather than a rolling stream.

A fresh example shows the runtime layer can drift into a breaking change, not just a behavioral one: Triton's 2.70.0 release removes Windows support and changes Python-client BF16 handling, the kind of bump a pinned major-version dependency doesn't protect against.

Drift can also be silently into a vulnerable version, not just a behaviorally different one: deptrust is a CLI that checks an agent's resolved package versions against known-vulnerability databases across a dozen ecosystems (npm, PyPI, crates.io, Go modules, and more) — a supply-chain-security check aimed specifically at what agents pull in as they write and modify dependency manifests, extending the drift discipline from "did behavior change" to "did the substrate become unsafe."

This is the obstacle that breaks an agent you already shipped, on a day you didn't deploy. You own the agent but rent the substrate, and its release cadence isn't yours — a framework patch can reintroduce a regression, an SDK bump can swap the executable, and a model deprecation can change behavior or pull the model entirely. The discipline is to treat the model, SDK, and serving runtime as pinned, version-controlled dependencies with a regression gate (see version pinning and agent benchmarks) — staged, tested upgrades, not a rolling "latest." Drift trades against freshness: the newest model or framework is also the one most likely to move under you.