Agent Budget Exhausted Halfway Through the Task
Your agent burns through its token or cost budget before finishing, leaving work incomplete. Here's how to diagnose spending and resume safely.
#AI coding
Articles tagged with #AI coding
Restored Agent Checkpoint Is Corrupted
Your agent resumes from a checkpoint but the state is garbled, missing fields, or internally inconsistent. Here's how to detect corruption and recover safely.
Cost Tracking Misses Sub-Agent Usage
Your pipeline's cost report shows $2 but the invoice says $18. Sub-agent token usage is not attributed to the parent run. Here's how to close the gap.
Cycle in Agent Call Graph Goes Undetected
Agents call each other in a loop that never terminates because the orchestrator has no cycle detection. Here's how to find cycles before they run forever.
Agent Handoff Loses Context Between Steps
Your multi-agent pipeline drops critical context at handoff boundaries. Learn to identify the gap and wire durable state in under an hour.
Agent Orchestrator Deadlocks Waiting on Each Other
Two agents block forever waiting for each other's output — a classic deadlock in async pipelines. Detect the cycle and break it in minutes.
Agent Output Leaks Secrets Into Downstream Logs
API keys, tokens, and passwords in agent output get written to logs and traces. Here's how to detect secret leakage and scrub it before it reaches storage.
Agent Output Not Machine-Parseable Downstream
Your agent wraps JSON in markdown or adds prose commentary, breaking the downstream parser. Here's how to enforce structured output reliably.
Agent Skipped a Pre-flight Check It Was Supposed to Run
Your agent proceeded without running required pre-flight checks, causing avoidable failures downstream. Here's how to enforce mandatory checks before execution.
Promotion Criteria Too Loose — Bad Output Slips Through
Your agent pipeline promotes flawed outputs because gate criteria are too broad or easy to satisfy. Here's how to tighten gates without over-blocking.
Prompt Templates Drift Between Agent Versions
Different versions of your agent use subtly different prompts, causing inconsistent output across the fleet. Here's how to version prompts and detect drift early.
One Agent's Rate Limit Cascades Into a Chain Failure
A single rate-limited agent stalls the entire pipeline as upstream agents queue and timeout. Learn to isolate rate limits and prevent cascade failures.
Flaky Tool Triggers an Agent Retry Storm
One unreliable tool call causes your agent to retry hundreds of times, exhausting budget and rate limits. Here's how to add backoff and circuit-breaking.
Shared Memory Corrupted by Overlapping Agent Writes
Two agents write to the same shared memory store simultaneously, producing garbled or inconsistent state. Here's how to detect and prevent overlapping writes.
Agent Skipped a Required Validation Step
Your agent bypassed a critical validation gate and pushed bad output downstream. Learn to enforce non-skippable checkpoints in any agent pipeline.
Agent State Desyncs After Restart
After a crash or restart, your agent believes the world is in a state that no longer matches reality. Here's how to detect drift and resync reliably.
Agent's Subprocess Orphaned After Agent Exits
Subprocesses launched by your agent keep running after the agent exits, consuming resources and causing side effects. Here's how to track and clean them up.
Task Routed to the Wrong Agent
Your orchestrator sends tasks to the wrong specialist agent, producing garbage output. Here's how to diagnose routing logic and fix misclassification.
Critical Tool Call Missing From the Agent Trace
Your agent trace shows a result but no corresponding tool call, hiding what actually happened. Here's how to find the gap and make traces complete.
Two Parallel Agents Edit the Same File
Parallel agents overwrite each other's changes, causing lost edits and broken code. Here's how to detect conflicts and enforce file-level locking.
AI Changelog Generation — From Commits to a Release Note Humans Read
Turn git log into release notes that aren't just a dump of commit messages.
AI-Assisted Database Migrations — Reversible, Backfilled, Tested
The three things AI gets wrong on migrations: reversibility, backfill, and the NOT NULL on a big table.
AI for Postmortems — Drafting Without Sanitizing the Lessons
Use AI to draft incident postmortems fast, without losing the uncomfortable truths.
AI for Merge Conflicts — When to Trust the Auto-Merge
AI can resolve most merge conflicts safely. Knowing which ones are NOT safe is the workflow.
AI On-Call Debugging — From Page to Fix Without Panic
How to use AI when you get paged at 3am — triage, hypothesize, calm down protocol, capture as runbook.
AI PR Descriptions — From Diff to Reviewable
Turn a git diff into a PR body a reviewer will actually read, with prompts that don't invent test plans for tests that don't exist.
Aider — Terminal-First AI Coding Without the Browser
Pip install, point it at a repo, and run the see-diff/accept loop from your shell — no IDE, no browser, no lock-in.
Claude Code MCP Servers — Wiring Up Real Tools
What MCP is, how to add a server with claude mcp add, three transport types, and the first servers worth installing.
Cursor Rules — Make .cursorrules Actually Earn Its Keep
Rules anatomy, when to use .cursor/rules/*.mdc vs the legacy .cursorrules, and how to write rules the model actually follows.
Windsurf — 30 Minutes to a Working AI-Coding Loop
Codeium's AI-first IDE walkthrough — install, configure, and run the first Cascade agent task without fighting the tool.
Cursor Cmd-K Inline Edit Spinner Never Returns
You hit Cmd-K, type the change, and the inline popup spins forever. No diff, no commit, no error. Usually network, model, or context. Walk the four root causes.
AI Added a Route That Bypasses Auth Middleware
AI registered a new endpoint outside the authenticated route group, exposing internal data without checks. Detect the gap and enforce auth as a default.
AI Invented a Wrong API Signature That Does Not Exist
AI confidently calls a method that does not exist — wrong arg order, made-up option name, fabricated return type. Diagnose by pinning to real source.
AI Generated Migration Works on Dev, Fails on Prod Schema
AI-written migration runs clean on an empty dev DB but blows up on prod with constraint or data errors. Fix with dev-prod parity and safety checks.
AI-Generated SQL Locks a Hot Table for Minutes
AI wrote a query or update that holds long locks on a production table, blocking reads and writes. Identify the pattern and rewrite it to be lock-friendly.
AI Keeps Using Deprecated Syntax Despite Lint Errors
AI generates code with deprecated APIs, fixes one lint error, then re-introduces the same pattern in the next file. Pin the rule and ground the prompt.
AI Uses npm Commands in a pnpm or Yarn Project
AI runs npm install and creates package-lock.json in a pnpm-only repo, breaking the lockfile invariant. Pin the manager and lock the rules.
Cursor Auto Model Picks Weak Model for Hard Task
Cursor's Auto model router routes a complex refactor to a small model, producing shallow or wrong edits. Diagnose routing signals and force the right model for hard tasks.
Cursor Python Venv Not Detected (Wrong Interpreter)
Cursor picks the system Python instead of your project venv, breaking imports, linting, and agent-generated code. Diagnose the interpreter picker, venv discovery, and shell PATH.
Cursor Settings Sync Overwrites Local Config on Other Machine
Cursor Settings Sync pushes outdated config to a second machine, clobbering local rules, keybindings, or MCP servers. Diagnose conflict direction, last-write-wins, and selective sync.
Cursor SSH Remote Connection Drops Mid-Edit
Cursor's SSH Remote session disconnects while you are editing, losing the agent context and unsaved buffer state. Diagnose the keepalive, network, and remote extension host.
Codex Committed Your Work to the Wrong Branch (or Straight to main)
Codex ran git commit on whatever branch was checked out — sometimes main, sometimes a leftover branch from the previous task. How to force a clean per-task branch.
Codex's Fix Passes Every Test but Breaks at Runtime
Codex's PR is green in CI — every test passes — but the app crashes in staging. Why agent fixes that target the test surface miss the runtime, and how to close the gap.
Every Codex Keyboard Shortcut, Explained
A complete Codex keyboard shortcut reference for the terminal TUI, desktop app, input editing, session control, mode switching, background tasks, and failure cases.
Codex Skills and /skills, Explained: Built-in, Custom, Plugin, Team-Shared
A complete Codex skills guide: what a skill is, how /skills discovery works, how to invoke skills, how to write SKILL.md, and how built-in, plugin, and custom skills differ.
AI Suggested a Stale Dependency
You asked for a library, it gave one that's been abandoned 4 years — training cutoff + no registry check.
AI-Generated Tests Pass But Feature Broken
All green, ship it, prod breaks — tests covered the happy path only, mocks shielded real branches.
AI Agents vs Autocomplete — When to Use Each
Autocomplete predicts the next 20 lines from local context; agents plan across files. Use this decision rule to map task shape to the right mode.
AI AdSense Readiness Review Tutorial
Use AI to pre-review your site against AdSense common rejection causes before applying.
AI Agent Code Review Workflow
Use Claude Code, Codex, or Cursor agents to pre-review your own PRs before sending them to humans.
AI Architecture Review Workflow
Use AI to challenge your architecture decisions before you commit weeks of code.
How to Use AI to Audit an Astro Content Site (Without Reading Every File)
A repeatable AI audit workflow for Astro content sites — catches broken slugs, missing translations, dead internal links, draft leakage, and config drift.
Use AI to Audit Astro Sites
Astro-specific audit prompts — sitemap, hreflang, output mode, RSS.
Use AI to Audit a React Native Project
AI can flag common React Native issues — re-renders, navigation, native modules.
Use AI to Find Broken Links Before Google Does
A monthly workflow to surface broken internal + external links using AI + simple tooling.
How to Use AI to Audit Category Pages (The Pages That Quietly Tank Sites)
Category pages are an under-audited SEO surface — this AI workflow catches duplicate intros, weak internal linking, missing descriptions, and bad pagination.
AI Codebase Tour Workflow — Onboarding to a New Repo
Use Claude Code, Cursor, or ChatGPT to give yourself a guided tour of an unfamiliar codebase.
AI Coding Context Management — What to Feed, What to Cut
AI coding quality is mostly a context problem. Here is how to feed agents the right context — and not the wrong one.
Plan a Pillar + Cluster Content Strategy with AI
Use AI to identify pillars, cluster topics, and content gaps for your niche.
AI Debugging Workflow — From Stack Trace to Fix
A reproducible workflow for debugging with AI that doesn't spiral into 20 wrong guesses.
AI Dependency Upgrade Workflow
Use Claude Code or Cursor to upgrade dependencies without breaking the build in subtle ways.
Use AI for Firebase Deploy Checks
Pre-flight checks via AI — firebase.json correctness, rewrites, function regions.
Using AI to Build Content Websites
Build a content site with AI assistance — content + structure + SEO.
AI Check Your Hreflang Setup (Bilingual / Multi-Locale)
Use AI to audit hreflang implementation — slug mismatches, missing pairs, wrong codes.
Batch-Generate Meta Descriptions with AI (Carefully)
Use AI to write meta descriptions at scale — without losing the quality that matters.
AI Migration Prompt Workflow — Framework / Language Migrations
A structured prompt workflow for big migrations: Vue 2→3, React class→hooks, JS→TS, Pages Router → App Router.
AI Coding in a Monorepo — Workflow That Doesn't Drown in Context
AI agents in monorepos either work great or melt down. Here is how to set them up to work.
AI Pre-Commit Review Workflow
A 60-second review pass with an AI agent before every commit. Catches the dumb stuff before it leaks.
AI Refactor Workflow: When It Works + When It Breaks
Refactoring with AI works when tests exist and scope is small.
AI Coding Rollback Workflow
Don't fear letting AI edit — fear losing the ability to undo.
How to Use AI to Triage Search Console Issues
A weekly 30-min AI workflow to triage Search Console: index coverage, Core Web Vitals, mobile usability, manual actions.
AI Spec-to-Code: Avoid the Half-Built-Feature Trap
Turn a written feature spec into shipped code without the half-built-feature trap.
How to Generate a Technical SEO Checklist with AI (Tailored to Your Stack)
Skip the generic 80-point SEO checklist. Generate one tailored to your stack (Astro / Next / Hugo / WordPress) with AI in 20 minutes.
AI Test Generation Workflow — Tests You Can Actually Trust
AI-generated tests often pass while testing nothing. Here is a workflow that gives you real coverage.
How to Use AI to Detect Thin Content (Before Google Does)
A 3-pass AI workflow that scores every page for thin content, surfaces the worst offenders, and proposes one of: expand, merge, noindex, delete.
App Project Audit Prompt Workflow
Use AI to audit a project for security, performance, and UX gotchas.
Bug Audit Prompt Workflow
Have AI surface likely bug spots in a module before they cause incidents.
ChatGPT for Coding — A Workflow That Survives Real Codebases
ChatGPT is not Claude Code, but used right it still ships real features. Here is how.
Claude Code Beginner Guide: Setup, First Refactor, and Mistakes to Avoid
What Claude Code is, how to install it, and how to ship your first refactor — the setup steps, prompt patterns, and common beginner mistakes.
Project Prompts for Claude Code
CLAUDE.md isn't just a readme — it's your agent's memory. Here's what to put in it.
Run a Full SEO Audit Using Claude Code
A reusable Claude Code prompt + workflow to run a real SEO audit on your content site.
Codex Beginner Guide: Sandboxed Cloud Tasks Without the Pitfalls (2026)
What Codex does, how it runs sandboxed cloud tasks, and when to use it. Setup, the spin-off workflow, and the mistakes that bite first-time users.
Codex for Code Review: Catch Bugs Before PR Review
Use Codex as a pre-review pair — catches issues before humans see them.
Use Codex to Review Your Sitemap
Use Codex to spot-check sitemap correctness, freshness, and coverage.
Coding Prompt Structure That Works
Goal + constraints + acceptance criteria. Skip any and the AI guesses.
Cursor Beginner Guide — From Install to First Real Edit
Cursor's real learning curve isn't install — it's knowing when to use Tab, Inline edit, Composer, Chat, or Agent. Four real edits, one per AI surface, in an hour.
Cursor Indexing — Make It Fast and Useful
A well-tuned index is the difference between Cursor "getting" your code and not.
How to Feed Project Reports to Coding Agents
A 30-page audit dumped into chat produces vague edits. Use this chunk-summarize-prioritize pattern to turn long reports into a brief the agent can execute.
Gemini for Coding Help — Where It Beats ChatGPT and Where It Doesn't
Gemini's coding has improved fast. Honest take on where it earns the click and where to switch to Claude or ChatGPT.
Git Commits with AI Agents
Commit messages, commit boundaries, and what NOT to let AI commit.
How to Review AI Diffs Efficiently
A 200-line diff isn't safe just because it compiles. Read it like a senior would.
Multi-Agent Coding Workflow
Use Cursor + Claude Code + Codex together - without conflicts.
Prevent Unsafe AI Edits: Guardrails That Actually Work
AI sometimes touches what it shouldn't. Guardrails that actually work.
SEO Audit Prompt Workflow
Use AI to audit your site's SEO basics - titles, descriptions, structured data, internal links.
AI Agent Loops Without Making Progress
Agent keeps trying the same fix or oscillates between two states. Break the loop fast.
Build Passes Locally but Fails After AI Edit in CI
Classic case: AI added a dep without committing the lockfile, or relied on a local Node version.
AI Code Edits Broke the Build — Recovery + Diagnosis
After an AI run, npm run build fails. Diagnose with diff review, type-checking, and selective revert.
AI Agent Overwrote .env / Environment Variables
Agent rewrote .env with placeholders or removed entries — recover and prevent.
AI Hallucinated a File That Doesn't Exist
Agent imports `src/utils/superhelper.ts` that was never created — common, fixable.
Merge Conflict After AI Edits — Resolution Guide
Your branch + AI edits conflict with main. Resolve without losing AI improvements.
AI Edits Broke package-lock / pnpm-lock Conflicts
Agent ran an install or removed deps and now your lockfile is at war with the team's.
AI Refactor Created Duplicate Files
Agent created `UserList2.tsx` next to the original or scattered duplicates. How to clean up safely.
AI Removed Working Logic During Refactor
AI "simplified" code and quietly removed a branch your users actually rely on.
Rolling Back AI Code Changes Safely
Three commands cover 99% of rollback scenarios: stash, restore, reflog.
AI Generated TypeScript Errors
AI produces code that doesn't type-check — usually wrong types, any-casts, or missing generic parameters.
How to Help Claude Code Understand Your Project: 7 Methods That Actually Work
Claude Code editing the wrong files, duplicating existing functions, missing key files? It's almost never the model — it's missing project context. Seven ways to make Claude Code actually read your codebase.
Codex Output Doesn't Match Project Structure
Codex creates files in the wrong place or with wrong conventions — constrain with AGENTS.md and explicit refs.
Cursor Cannot Apply Edits — Top Causes
Cursor shows edits but Apply does nothing or fails silently.
Cursor Missed the Full Project Context
Cursor edits only what it can "see" — fix by adjusting indexing scope and explicit @file references.
Cursor Keeps Reading the Wrong Files
Cursor pulls in old / vendored / generated files into context. Fix the index and ignore lists.
Recovering an Old File Version After Commit
You committed over what you wanted to keep. Three paths back. Usual causes: accidental commit overwrite; hard reset moved head. Start with: `git log -p -- <file>` to see the file's history.
How to Inspect AI-Generated Diffs Before Accepting
A 200-line AI diff isn't safe just because it compiles. Read it in the right order.
Multiple AI Agents Created Conflicts
Two agents (Cursor + Claude Code) edited the same file. Merge cleanly, don't panic.
Accessibility Audit Prompts for WCAG Frontend Reviews
12 prompts that walk a page like a keyboard-only and screen-reader user — semantics, focus, ARIA widgets, contrast, modal stacks, live regions, WCAG mapping.
API Contract Review Prompts for REST and GraphQL
15 prompts that pressure-test REST and GraphQL contracts before they ship — naming, status codes, error model, breaking changes, pagination, N+1, field-level auth.
Bug Audit Prompts: Hunt Hidden Bugs Before Prod
13 prompts to systematically hunt hidden bugs in code before they ship.
Claude Code Execution Prompts: Plan, Build, Verify
12 prompts to brief Claude Code (or Codex) on real engineering tasks — scoped features, surgical bug fixes, migrations, refactors, TDD, perf work, debugging, code review.
Code Review Prompts: Beyond "Looks Good to Me"
13 prompts for AI code review that surfaces real issues — bugs, security, performance, readability, tests.
Database Schema Review Prompts for Safe Migrations
12 prompts that walk a schema like a future migrator — normalization, indexes, FK cascades, nullability, soft-delete, JSON columns, migration safety, ORM diff.
Deployment Check Prompts: Pre-Launch and Post-Deploy Checklists
12 prompts to pre-flight your deployment — env vars, DNS, cache, observability, rollback drill, post-launch monitoring, security headers, status page comms.
Performance Optimization Prompts: 13 Templates for Real Speed Wins
13 prompts for performance work that actually moves a number — measure-first bottleneck ID, LCP / CLS / N+1 / TTFB / bundle / cache fixes, before-after Web Vitals diffs, and a premature-optimization detector.
React Component Refactor Prompts for Cleaner Components
13 prompts to refactor React components — extract hooks, lift state, split server vs client, memoize correctly, kill prop-drilling.
README Generation Prompts: Get Reader to "Hello World"
12 prompts for READMEs that get a new reader from git clone to a working install in 60 seconds — skeleton, quickstart, install troubleshooting, examples, API reference, and audit.
Refactor Prompts: 18 Templates for Safer AI Refactors
18 copy-ready prompts to plan, execute, and verify AI refactors without breaking behavior — module extraction, renames, async migrations, and god-function splits.
Security Audit Prompts: AppSec for Indie Devs
12 prompts to AI-audit your code for OWASP-grade security issues — auth, authz, secrets, deps, file uploads, CORS, PII logging — without hiring a pentester.
SEO Audit Prompts for Content Sites
12 prompts to AI-audit a content site — metadata, internal links, hreflang, thin content, canonical, schema, cannibalization, Core Web Vitals — with measurable thresholds and fixes.
Test Generation Prompts (Integration / E2E / Snapshot): 13 Templates
13 prompts for integration, E2E, snapshot, and contract tests — for unit-test prompts specifically, see the unit-test article. Tests that catch real bugs, not noise.
TypeScript Error Diagnosis Prompts for Cryptic TS Errors
14 prompts to diagnose TypeScript errors fast — generics, conditional types, narrowing, module resolution, declaration files.
Claude Code vs Cursor: Which Task Belongs to Which
Claude Code is a terminal agent; Cursor is an in-editor copilot. Task fit, debug loop, and rollback cost differ. A decision sheet.
Cursor for Beginners: 30 Minutes to a Working Loop (2026)
From install to your first useful edit in 30 minutes: Inline vs Composer vs Chat, codebase indexing, @Files context injection, Apply, and rollback.
Cursor Stuck Indexing: Make It Skip node_modules and Build Output
Cursor stuck on Indexing? 90% of the time it's trying to index node_modules, dist, .next, or cache directories. Here's the .cursorignore template and a clean rebuild flow.
Codex vs Claude Code: Which AI Coding Agent Should You Use?
OpenAI Codex and Anthropic Claude Code are both AI coding agents — but they live in different places and solve slightly different problems. Here's a clear comparison.
What Is Claude Code? Anthropic's Official Terminal AI Explained
Claude Code is Anthropic's command-line AI tool. This guide explains what it does, how it differs from ChatGPT and Cursor, and how to get productive in the first ten minutes.