Everyone's raving about iboss's free tier like it's manna from heaven for developers. Let's cut through the hype. It's not a generous offering; it's a classic vendor lock-in play disguised as a free lunch. You get just enough functionality to become dependent on their workflow, only to hit a hard wall the moment you need something real.
I tried integrating it into a standard CI pipeline for a Python microservice. The free tier's API rate limits are a joke—you can't even run a full test suite without getting throttled. Their documentation conveniently buries this detail until you've already wired it into your `pytest` hooks. You end up with something like this:
```python
# In your conftest.py, feeling clever
def pytest_sessionstart(session):
response = iboss_api.scan_code(session.items)
# Surprise! Free tier limit: 10 scans/hour
if response.status_code == 429:
pytest.skip("iboss rate limit hit, tests unreliable")
```
Suddenly, your test results are non-deterministic, based on whether you've made too many commits that hour. The "limited view" they give you is engineered to feel just barely useful, until it isn't. You start seeing "upgrade for more detailed analysis" prompts on what look like trivial security nudges, pushing you toward the paid plan to get any actionable intelligence.
The real trap is the context shift. You begin structuring your code and commits around their free tier's constraints, not best practices. Then, when you outgrow it, migrating away is painful because their analysis format is proprietary. You're not just paying for a service; you're paying to escape the workflow they let you build for free. Clever, and deeply annoying.
prove it to me