Practical notes: I built an AI Doc-Fixing Agent. Here’s every way it fooled

Use this as an operator-facing rebuild of the ideas in “I built an AI Doc-Fixing Agent. Here’s every way it fooled itself first.”: clear stages, ordered code slots, and recovery notes that survive a handoff. The Overview stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph.

- print(tqdm.__version__, sys.version, sys.platform)
+ print(tqdm.__version__, sys.version, sys.platform)

How it fits together

For the How it fits together stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness.

Round 1: tqdm — the fix that fixed nothing

For the Round 1 tqdm the stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness.

Found 9 Markdown file(s). Starting audit...

Auditing: .github\ISSUE_TEMPLATE\bug.md
  Generating repair for snippet #1
  Verifying generated repair for snippet #1
  Publishing verified repair(s)...
  VERIFIED REPAIR: Branch published

- print(tqdm.__version__, sys.version, sys.platform)
+ print(tqdm.__version__, sys.version, sys.platform)

Round 2: click — the failure that repeated itself six times

For the Round 2 click the stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Put human approval on edges that spend money or change production data. Compile-time wiring does not equal business completeness. For the Round 2 click the stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph.

Auditing: docs\advanced.md
  Generating repair for snippet #2
  Verifying generated repair for snippet #2
  [Reflection Loop 2/3] Revising repair for snippet #2
  ...
  [Reflection Loop 3/3] Revising repair for snippet #3
  NOT PUBLISHED: Generated repair failed verification:
    ModuleNotFoundError: No module named 'click'

Finding the root cause — with Docker, manually

When working through the Finding the root cause stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.

command.extend(["-v", f"{project_dir}:/workspace", "-w", "/workspace", "-e", "PYTHONPATH=/workspace"])

docker run --rm -v "C:\temp\click_test:/workspace" -w /workspace -e PYTHONPATH=/workspace python:3.10-slim python -c "import click"
# ModuleNotFoundError: No module named 'click'

docker run --rm -v "C:\temp\click_test:/workspace" -w /workspace -e PYTHONPATH=/workspace/src python:3.10-slim python -c "import click; print(click.__version__)"
# import works... but:
# importlib.metadata.PackageNotFoundError: No package metadata was found for click

The fix: stop guessing, let pip do its job

When working through the The fix stop guessing stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.

Re-running click with the fix in place

When working through the Re-running click with the stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node. When working through the Re-running click with the stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph.

@cli.result_callback()
def process_pipeline(processors, fin):
- iterator = (x.rstrip("\r\n") for x in input)
+ iterator = (x.rstrip("\r\n") for x in fin)

What you’d tell someone testing an AI agent like this

The What you d tell stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.

Where it stands now

The Where it stands now stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Keep graph state flat and typed. Nested blobs hide which node wrote which field and break resume after interrupts.

Operational checklist

When working through the Operational checklist stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest.

Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.

Pin dependency versions and record the image digest that ran the demo. Reproducibility beats tribal knowledge.

Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph.

Checkpoint after expensive steps. Resume should not re-bill the same LLM call when an operator retries a later node.

Before promoting the stack, freeze versions, capture a golden transcript for the critical path, and confirm rollback steps. Shared environments need rate limits, tenancy checks, and a clear owner for secret rotation. Prefer boring reliability over clever one-off demos.

Batch note for 21079fcb54b5: keep provider keys out of the repo, set a per-session token ceiling, and store transcripts next to the eval fixtures so later model swaps stay comparable.

When working through the hardening note 0 stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion.

Hardening detail 0/814: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

The hardening note 1 stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph.

Hardening detail 1/814: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.