This demo shows VALP’s narrow value: a task is not done just because an agent or runtime says it is done. Expected evidence must exist.
The demo uses the bundled Manual Mode example and does not launch a live runtime.
git clone https://github.com/wcqxgjy6d8-pixel/Visible-Agent-Loop-Protocol.git
cd Visible-Agent-Loop-Protocol
python -m pip install -r requirements-dev.txt
bin/valp audit examples/minimal-task
Expected result:
VALP audit: PASS
Summary: pass=13 warn=0 fail=0 skip=6
Copy the task to a temporary directory, then remove the review evidence that the receipt ledger expects.
demo_dir="$(mktemp -d /tmp/valp-demo.XXXXXX)"
cp -R examples/minimal-task "$demo_dir/minimal-task"
mv "$demo_dir/minimal-task/agents/manual-reviewer/review.md" \
"$demo_dir/minimal-task/agents/manual-reviewer/review.md.bak"
bin/valp audit "$demo_dir/minimal-task"
Expected result:
VALP audit: FAIL
Summary: pass=12 warn=0 fail=1 skip=6
[FAIL] expected_evidence: Expected evidence exists
Missing expected evidence: agents/manual-reviewer/review.md
mv "$demo_dir/minimal-task/agents/manual-reviewer/review.md.bak" \
"$demo_dir/minimal-task/agents/manual-reviewer/review.md"
bin/valp audit "$demo_dir/minimal-task"
Expected result:
VALP audit: PASS
Summary: pass=13 warn=0 fail=0 skip=6
This proves the evidence discipline, not live automation. The repository can detect that a task folder is missing expected proof, even when the latest receipt says the manual result was attested.
Full Mode claims require stronger runtime proof: a real runtime submission receipt, expected evidence, review/verification status, approval resolution when needed, and final synthesis.