All
Articles 129,974Blog Posts 134,416Tech Tutorials 33,616Research Papers 25,164News 18,387
⚡ AI Lessons

Dev.to · Jeremy Longshore
3d ago
Copying Files Is Not Installing
A marketplace install copies a plugin's files into place but never runs its installer. Here is how that bricked local mode and the dependency-free bootstrap tha

Dev.to · Jeremy Longshore
5d ago
Producer Fallback: When Claude Hits the Weekly Limit, the Pipeline Still Ships
A weekly LLM limit is a producer outage, not a reason to miss the daily post. Grok fallback + land path kept startaitools shipping.

Dev.to · Jeremy Longshore
5d ago
Exit 0 Is Not Success: Automation Assurance That Verifies Outcomes
A green exit code can hide a dead job. Intent-OS automation assurance treats outcome verification as the success criterion, not exit 0.

Dev.to · Jeremy Longshore
6d ago
Empty Is Not Clean: Five Fail-Open Bugs in an AI Agent
One fail-open bug, five disguises: empty args, an empty log, a missing property, a sheared chain, an unset scope. When there's nothing to check, fail closed.

Dev.to · Jeremy Longshore
1w ago
Liveness Without Health Is Theater
A heartbeat that fires on every run proves a job ran, never that it succeeded — the fix is two markers plus meta-monitoring the watchers.

Dev.to · Jeremy Longshore
1w ago
Adversarial Review: The Six Lenses That Halted a Rollout
A six-lens adversarial review checked a team knowledge system against live state, broke three shipped assumptions, and gated 18 risks to halt the rollout.

Dev.to · Jeremy Longshore
🔍 RAG & Vector Search
⚡ AI Lesson
2w ago
The Relevance Score That Broke Our Cite-or-Refuse Gate
A retrieval test asserted every relevance score sits in [0,1] — it failed on a live -1.83 silently corrupting a cite-or-refuse RAG safety gate.

Dev.to · Jeremy Longshore
☁️ DevOps & Cloud
⚡ AI Lesson
2w ago
Run the Readiness Audit Before You Flip DNS
An adversarial readiness audit found a payment-path bug that would have charged every customer, then returned HTTP 500 — before the irreversible DNS flip.

Dev.to · Jeremy Longshore
2w ago
Gate the Statement, Not the Tool Name
When one MCP tool carries every SQL verb, allowlisting tool names is theater. The safety boundary has to read the statement — here's how that gate was built.

Dev.to · Jeremy Longshore
📐 ML Fundamentals
⚡ AI Lesson
2w ago
Coverage Said 69%, Mutation Testing Said 25%
A repo at 69% line coverage scored 24.88% on mutation testing—and the rules engine that touches user email scored 0.00%. Coverage said fine; Stryker didn't.

Dev.to · Jeremy Longshore
☁️ DevOps & Cloud
⚡ AI Lesson
3w ago
Green CI Proves Nothing: Why Your Tests Gate Zero Calls
CI dogfood for AI-agent governance went green while gating zero tool calls. Here's why a passing test proving nothing is worse than a red one.

Dev.to · Jeremy Longshore
⚡ AI Lesson
3w ago
MCP Server Auth: The API Is the Real Boundary
Per-user tokens, a server-side write gate, and a separate access log — why an MCP server's client-side tool gate is UX, not a security boundary.

Dev.to · Jeremy Longshore
☁️ DevOps & Cloud
⚡ AI Lesson
3w ago
When --cap-drop ALL Broke the Gate Socket
Hardening a container hid a permission bug: --cap-drop ALL stripped CAP_DAC_OVERRIDE, and a gate socket silently stopped governing every tool call.

Dev.to · Jeremy Longshore
⚡ AI Lesson
1mo ago
Honor the Gate When the Verdict Is Inconvenient
A quality gate only matters if you honor its verdict. How pre-registration and honest-gate culture stopped two teams from faking green or rationalizing a STOP.

Dev.to · Jeremy Longshore
🤖 AI Agents & Automation
⚡ AI Lesson
1mo ago
Human-in-the-Loop Is a Delivery Guarantee, Not a UI Feature
Human-in-the-loop agent delivery is exactly-once, fail-closed. Two repos shipped the same four-move discipline the same day — convergence, not coincidence.

Dev.to · Jeremy Longshore
☁️ DevOps & Cloud
⚡ AI Lesson
1mo ago
Vite Dev Server in Production: The 871-Byte Tell
scorecardecho.com shipped the Vite dev server to every visitor. Three signals catch it in a minute; a multi-stage Dockerfile fixes it for any SPA container.

Dev.to · Jeremy Longshore
1mo ago
The Unicode Layer Your Validator Can't See
Schema validation can't see invisible Unicode. A stdlib-only CI gate that catches tag-char injection, Trojan Source bidi overrides, and homoglyph attacks.

Dev.to · Jeremy Longshore
☁️ DevOps & Cloud
⚡ AI Lesson
1mo ago
Self-Expiring Report-Only CI Gates: From Advisory to Enforced
How a meta-gate enforces deadline-driven CI hardening without freezing contributors — one logical concern per PR, permanent blocking by design.

Dev.to · Jeremy Longshore
🔐 Cybersecurity
⚡ AI Lesson
1mo ago
Safety Model First: 16-Tool Ops MCP, One Day
Design a 7-point safety model before writing tools. How server-ops-mcp shipped 16 tools, 40 tests, and v0.1.0 in a single day.

Dev.to · Jeremy Longshore
2mo ago
A v1.0 Is a Gate, Not a Tag
Why release gates should accept GO with conditions, not binary GO/NO-GO. How ICO v1.0.0 shipped with documented gaps and a same-day v1.0.1.

Dev.to · Jeremy Longshore
⚡ AI Lesson
2mo ago
Honest Perf Benchmarks for a Paid-API Compiler
Four PRs, three releases, and a benchmark suite that won't lie to you: seeded-RNG corpora, double-gated Claude scenarios, and skipped-but-recorded records.

Dev.to · Jeremy Longshore
🧠 Large Language Models
⚡ AI Lesson
2mo ago
Deterministic First, LLM Second: An Advisory CI Pre-Screen
Wire an LLM into CI without veto power: the deterministic classifier is the product, the LLM is advisory polish, and the never-block contract retires the old sy

Dev.to · Jeremy Longshore
2mo ago
Transitive CVE Clearance: The Dual-Layer Pattern
How v0.9.1 cleared 6 high-severity transitive CVEs in axios + fast-uri — and why the dep bump needed top-level overrides to actually stick.

Dev.to · Jeremy Longshore
☁️ DevOps & Cloud
⚡ AI Lesson
2mo ago
Two False-Positive Fixes, Same Root Cause
Two separate false-positive alerts, same root cause: monitoring conjoined liveness with conditional behavior. Separating health signals quiets noise under load.
DeepCamp AI