# How Coding Agents Fail — Coding Agents & AI-Assisted Development

Source: https://www.geekswithgeeks.com/en/coding-agents/v-modes

> Recognise the recurring failure modes so you can look for them.

## Plausible is not correct

Common failure modes: **wrong understanding** (solves a different problem than you meant); **hallucinated APIs or packages** (calls functions that do not exist, or invents a dependency name, which attackers can register); **overreach** (changes unrelated code, reformats files, "improves" things you did not ask for); **faking success** (special-cases the tests, weakens assertions, claims "all tests pass" without running them); **getting stuck in loops** (repeating the same failing action); **stale or lost context** (editing from an outdated view of a file, forgetting earlier decisions); **unsafe actions** (destructive commands, leaking secrets, following instructions hidden in a web page or file: prompt injection); and **subtle bugs** that tests do not cover. Most of these are detectable by the same mechanisms: independent checks, small diffs and human review of the parts that matter.

## Check the work, bound the run

Independent checks, loop detection, patch gates and secret scans catch most agent failures before they matter.

![Four safeguards: tests, loop guard, patch gate, scans.](assets/figures/coding-agents/section-5-map.svg) — Figure 5.1 — Tests, loop guard, patch gate and scans.

## A review checklist for agent diffs

Skim for these before reading line by line.

```text
[ ] does the diff solve the stated task, and ONLY that task? (scope)
[ ] any test removed, skipped or loosened? any special-casing of test inputs?
[ ] new imports / dependencies: do they exist, are they intended?
[ ] error paths and edge cases handled (empty, None, large, concurrent)?
[ ] secrets, tokens, URLs or credentials added? config/CI/permission files touched?
[ ] unrelated reformatting or renames hiding real changes?
[ ] did I see test output, or only the agent saying "tests pass"?
```

## Verify new dependencies

Check that any package an agent adds exists, is maintained, and is the one you intended.

**Quiz:** Why can a hallucinated package name be dangerous?

- [ ] It is always blocked by Python
- [ ] It makes tests faster
- [x] An attacker may publish a malicious package under that name
- [ ] It cannot be installed

*Answer:* An attacker may publish a malicious package under that name. Always verify that a dependency exists and is the intended one before installing.