agnes-the-ai-analyst/tests/test_store_put_atomic.py
Vojtech d6ad08f107
Flea-market upload guardrails + soft delete + JOIN-based admin queue (#233)
* feat(store): flea-market upload guardrails + soft delete + JOIN-based admin queue

Adds an end-to-end guardrails pipeline for store uploads (manifest +
static-security + LLM review), persists blocked bundles for forensics,
introduces soft-delete (Archive) semantics, consolidates the legacy
/store/{id} surface into /marketplace/flea/{id}, and reworks the admin
queue so lifecycle filters read live entity visibility via LEFT JOIN
rather than a denormalized submission column.

Schema v29 → v35:
  * v29 store_submissions table + store_entities.visibility_status
  * v30 file_size, bundle_sha256, bundle_purged_at on submissions
  * v31 reshape store_submissions (drop legacy unique on entity_id)
  * v32 store_entities.archived_at/by + 'archived' visibility value
  * v33 drop store_submissions.retry_count (unused)
  * v34 ensure idx_store_submissions_entity exists post column-drop
  * v35 broaden visibility_status enum + JOIN architecture cutover

Pipeline (src/store_guardrails/):
  * Inline checks: manifest_check, static_scan, quality_check
  * LLM review configurable haiku|sonnet|opus (default haiku)
  * BackgroundTasks-driven async path with structured-output JSON
  * Per-submitter daily quota (default 50)
  * 30-day TTL purge job (POST /api/admin/run-blocked-purge)
  * Bundle SHA256 + size persisted; sha256 survives purge for forensics

Visibility model:
  * pending | approved | hidden | archived
  * _enforce_visibility returns 404 (no leak) for non-owner non-admin
  * Owner sees own non-approved entries via include_owner_id widening
  * Install refused with 409 entity_not_approved when not approved

Soft-delete (DELETE /api/store/entities/{id}):
  * Default = soft (visibility_status='archived'); existing installs
    keep getting served the bundle so users don't lose the plugin
  * ?hard=true admin-only: drops bundle + cascades user_store_installs
  * Hard-delete preserves entity_id on submission as tombstone so
    audit_log linkage survives for the activity timeline

Admin queue lifecycle (the JOIN refactor):
  * Verdict (store_submissions.status) is immutable forensic record
  * Lifecycle (store_entities.visibility_status) is live state
  * /admin/store/submissions Archived chip translates to
    `e.visibility_status='archived'` via LEFT JOIN — any path that
    flips visibility surfaces in the queue immediately
  * Detail page renders Status (verdict) and Entity lifecycle side by
    side so admins see "approved at review, now archived" at a glance

URL consolidation:
  * /store/{id} deleted (no redirect, stale bookmarks 404)
  * /marketplace/flea/{id} is the canonical detail surface
  * Three in-tree callers (upload-success, my-stack card, store
    listing card) updated to point at the new URL
  * Quarantine banner extracted to _quarantine_banner.html partial,
    self-guarded, included from both flea detail templates
  * Banner JS auto-refreshes when the verdict lands by polling
    /api/marketplace/flea/{id}/detail (visibility_status +
    submission_status — the latter is needed because blocked_llm
    keeps the entity at visibility_status='pending')

Audit log resource format:
  * runner.py emits prefixed `store_submission:{id}` (post-fix)
  * Detail-page timeline query handles three patterns: prefixed
    submission, helper-emitted `store_entity:{sub_id}`, and bare-id
    legacy rows — all surface in the activity timeline

UX fixes:
  * Owner sees Under review / Quarantined / Hidden banner with status
  * Install button gray-disabled (not blue) when non-approved
  * Owner cannot delete quarantined entries (403); admin can
  * Admin queue: filter chips, sortable columns, paging, page-size
  * Auto-refresh queue every 5s while pending rows are visible
  * Store upload page file picker no longer opens twice (label →
    input default action collided with explicit JS handler)

Tests: 168 passed across the guardrails suites (admin submissions,
store API, inline / LLM / purge guardrails, store repositories,
marketplace filter, schema version). New regression coverage
includes: archive surfaces via JOIN even when API path is bypassed;
deleted submission renders activity timeline (tombstone); flea
detail surfaces submission_status only for owner/admin; detail page
renders Entity lifecycle row; audit log resource format covers both
helper and runner paths.

* fix(store-guardrails): PR #233 follow-up — prompt injection, atomic PUT, BG race, schema, reaper, sort whitelist

Addresses 9 of the 23 findings from the PR #233 review (spec at
docs/superpowers/specs/2026-05-09-pr233-guardrails-fixes-spec.md).
Merge-gate items #1-#6 plus high-value mediums #7, #9-#12, #23.
Architectural items (#8 enum split, #14 factory) and pure
maintainability (#15-#22) deferred to follow-ups.

Security:
* #1 prompt injection — SYSTEM_PROMPT now passed via the SDK's
  dedicated system= parameter; bundle wrapped in <bundle>...</bundle>
  sentinels declared data-only by the system prompt; literal
  sentinel strings in user content are escaped so an adversarial
  README can't forge a close tag.
* #6 static scan honesty — module docstring + admin copy + docs
  declare static scan as signal not gate; .md/.txt/.rst/.html/.json/
  .yaml/.yml/.toml skipped to avoid false positives on prose.
  AST mode for Python deferred (separate flag, FP comparison work).

Correctness:
* #2 PUT atomicity — bundles bake into plugin.staging-<rand>/
  alongside live, atomic-rename on success; failed checks leave
  live tree byte-for-byte intact.
* #3 BG-task race — set_visibility_if_pending guards verdict flips
  to the (pending, hidden) review window; admin archives during
  review survive; skipped flips audit-logged.
* #4 v35 NOT NULL/DEFAULT — schema v35→v36 re-applies them on
  store_entities.visibility_status. CHECK constraint enforced
  application-side (DuckDB ADD CHECK on existing column unsupported).
* #7 stuck-review reaper — reap_stuck_llm_reviews flips pending_llm
  rows older than guardrails.stuck_review_grace_seconds (default
  1800) to review_error. Scheduler runs every 15 min via new
  /api/admin/run-reap-stuck-reviews. Set knob to 0 to disable.
* #9 quota counter — count_blocked_for_submitter_since now counts
  blocked_inline + blocked_llm + review_error so a submitter
  triggering only LLM-blocked verdicts is bounded.
* #10 missing risk_level — surfaces as review_error with
  error='missing_risk_level' instead of silently defaulting to
  'medium' (which looked like a model-decided block).
* #11 archived_at clear — set_visibility nulls archived_at +
  archived_by when transitioning out of 'archived' so a future
  read doesn't show stale archive forensics on an approved row.

Maintainability:
* #12 FSM doc comment — accurate insert/transition/lifecycle
  description in src/db.py near store_submissions schema.
* #23 sort-key whitelist — admin queue rejects unknown sort keys
  with 400 invalid_sort_key; substring-replace footgun removed.

Deferred (separate PRs):
* #5 quota race — proper fix requires asyncio.Lock spanning the
  full pipeline; threading.Lock blocks event loop, DuckDB MVCC
  doesn't help. API-level slowapi bounds worst case for now.
* #6 part 3 (AST static scan), #8 (enum split), #13 (import
  bundle docs), #14 (factory consolidation), #15-#22 (maint).

Tests:
* New: tests/test_store_guardrails_prompt_injection.py (corpus +
  trust-boundary invariants), tests/test_store_put_atomic.py,
  tests/test_store_guardrails_reaper.py.
* Extended: test_store_guardrails_llm.py (system param, missing
  risk_level, BG race), test_admin_store_submissions.py (quota
  counter widening, sort whitelist 400), test_store_repositories.py
  (un-archive metadata clear), test_db_schema_version.py (v36).
* Full suite: 3738 passed; 17 pre-existing baseline failures
  unchanged (db migration tests, cli binary rename, catalog export,
  user mgmt v5 backfill — confirmed by stash + rerun on clean tree).
2026-05-09 17:32:53 +04:00

223 lines
8.2 KiB
Python

"""PUT /api/store/entities/{id} atomicity (#2 from PR #233 review).
Pre-fix: the bake wrote into the live `${DATA_DIR}/store/<id>/plugin/`
path BEFORE running guardrail checks. A concurrent GET during the
window saw partial / unverified content, and a failed check left the
on-disk tree in a partially-overwritten state until the rollback
copytree finished.
Post-fix: bake into a sibling `plugin.staging-<rand>/` dir, run checks
there, then atomic rename onto the live path. Failed checks leave the
live tree byte-for-byte intact.
"""
from __future__ import annotations
import hashlib
import io
import zipfile
from pathlib import Path
import pytest
from argon2 import PasswordHasher
from fastapi.testclient import TestClient
from src.db import close_system_db, get_system_db
from src.repositories.users import UserRepository
@pytest.fixture
def web_client(tmp_path, monkeypatch):
monkeypatch.setenv("DATA_DIR", str(tmp_path))
monkeypatch.setenv("TESTING", "1")
monkeypatch.setenv("JWT_SECRET_KEY", "test-secret-key-min-32-characters!!")
(tmp_path / "state").mkdir()
(tmp_path / "analytics").mkdir()
(tmp_path / "extracts").mkdir()
close_system_db()
from app.main import create_app
app = create_app()
yield TestClient(app)
close_system_db()
def _create_user(client, email, password="UserPass1!"):
ph = PasswordHasher()
conn = get_system_db()
user_id = email.split("@")[0]
UserRepository(conn).create(
id=user_id, email=email, name=user_id, password_hash=ph.hash(password),
)
conn.close()
r = client.post("/auth/token", json={"email": email, "password": password})
assert r.status_code == 200, r.text
return user_id, {"access_token": r.json()["access_token"]}
def _make_skill_zip(skill_name: str, body: str) -> bytes:
buf = io.BytesIO()
with zipfile.ZipFile(buf, "w") as zf:
zf.writestr(
f"{skill_name}/SKILL.md",
f"---\nname: {skill_name}\ndescription: A clean test skill for atomic-PUT testing.\n---\n\n"
+ body,
)
return buf.getvalue()
def _make_evil_zip(skill_name: str) -> bytes:
"""A skill containing a static-security violation (eval) — fails
inline checks during PUT, so the live tree must NOT be touched."""
buf = io.BytesIO()
with zipfile.ZipFile(buf, "w") as zf:
zf.writestr(
f"{skill_name}/SKILL.md",
f"---\nname: {skill_name}\ndescription: Updated body content.\n---\n\nBody. " * 30,
)
zf.writestr(f"{skill_name}/run.sh", "#!/bin/sh\neval $1\n")
return buf.getvalue()
def _hash_tree(root: Path) -> str:
"""Stable digest of the on-disk plugin tree (path + content)."""
h = hashlib.sha256()
for p in sorted(root.rglob("*")):
if not p.is_file():
continue
rel = p.relative_to(root).as_posix().encode()
h.update(rel + b"\0" + p.read_bytes() + b"\0")
return h.hexdigest()
def _plugin_dir_for(entity_id: str) -> Path:
"""Mirror app/api/store.py:_plugin_dir without importing private."""
from app.utils import get_store_dir
return Path(get_store_dir()) / entity_id / "plugin"
class TestPutAtomicity:
def test_failed_inline_check_leaves_live_tree_intact(self, web_client):
"""The live `plugin/` tree must be byte-for-byte identical
before and after a PUT whose bundle fails inline checks."""
owner_id, owner_cookies = _create_user(web_client, "ownerA@x.com")
clean_zip = _make_skill_zip("atomic-skill", "Clean body. " * 30)
c = web_client.post(
"/api/store/entities",
files={"file": ("s.zip", clean_zip, "application/zip")},
data={"type": "skill"}, cookies=owner_cookies,
)
assert c.status_code == 201, c.text
eid = c.json()["id"]
plugin_dir = _plugin_dir_for(eid)
before_hash = _hash_tree(plugin_dir)
assert before_hash, "expected non-empty plugin tree"
# PUT with a bundle that will fail static_security (contains eval).
evil_zip = _make_evil_zip("atomic-skill")
u = web_client.put(
f"/api/store/entities/{eid}",
files={"file": ("evil.zip", evil_zip, "application/zip")},
cookies=owner_cookies,
)
# Inline-blocked uploads return 422 with a structured detail.
assert u.status_code == 422, u.text
assert u.json()["detail"]["code"] == "submission_blocked"
after_hash = _hash_tree(plugin_dir)
assert after_hash == before_hash, (
"live plugin tree changed after a failed-check PUT — "
"atomic-rename invariant broken"
)
# Sibling staging dirs must not be left behind.
entity_root = plugin_dir.parent
leftovers = [
p for p in entity_root.iterdir()
if p.name.startswith("plugin.staging-")
or p.name.startswith("plugin.backup-")
]
assert not leftovers, (
f"staging/backup dirs leaked on disk: {leftovers}"
)
def test_successful_put_atomically_replaces_tree(self, web_client):
"""Successful PUT swaps the live tree to the new bundle without
leaving a staging dir behind."""
owner_id, owner_cookies = _create_user(web_client, "ownerB@x.com")
v1 = _make_skill_zip("swap-skill", "First body. " * 30)
c = web_client.post(
"/api/store/entities",
files={"file": ("v1.zip", v1, "application/zip")},
data={"type": "skill"}, cookies=owner_cookies,
)
assert c.status_code == 201, c.text
eid = c.json()["id"]
plugin_dir = _plugin_dir_for(eid)
before_hash = _hash_tree(plugin_dir)
v2 = _make_skill_zip("swap-skill", "Second different body. " * 30)
u = web_client.put(
f"/api/store/entities/{eid}",
files={"file": ("v2.zip", v2, "application/zip")},
cookies=owner_cookies,
)
assert u.status_code == 200, u.text
after_hash = _hash_tree(plugin_dir)
assert after_hash != before_hash, "PUT didn't change live tree"
entity_root = plugin_dir.parent
leftovers = [
p for p in entity_root.iterdir()
if p.name.startswith("plugin.staging-")
or p.name.startswith("plugin.backup-")
]
assert not leftovers, (
f"staging/backup dirs leaked on disk after success: {leftovers}"
)
def test_inline_check_failure_during_put_does_not_pollute_tree(
self, web_client, monkeypatch,
):
"""Force a check failure mid-bake by monkey-patching
run_inline_checks. Live tree must still be intact."""
from src.store_guardrails.runner import InlineResult
owner_id, owner_cookies = _create_user(web_client, "ownerC@x.com")
clean_zip = _make_skill_zip("monkey-skill", "Body. " * 30)
c = web_client.post(
"/api/store/entities",
files={"file": ("v1.zip", clean_zip, "application/zip")},
data={"type": "skill"}, cookies=owner_cookies,
)
assert c.status_code == 201, c.text
eid = c.json()["id"]
plugin_dir = _plugin_dir_for(eid)
before_hash = _hash_tree(plugin_dir)
# Force the PUT path to see a failed inline result without
# actually relying on a static_security regex match.
def fake_inline(*args, **kwargs):
return InlineResult(
manifest={"status": "fail", "issues": ["forced"]},
static_security={"status": "pass", "findings": []},
quality={"status": "pass", "issues": [],
"template_placeholders": 0,
"template_recommendation": None},
)
monkeypatch.setattr(
"app.api.store.run_inline_checks", fake_inline,
)
v2 = _make_skill_zip("monkey-skill", "Different. " * 30)
u = web_client.put(
f"/api/store/entities/{eid}",
files={"file": ("v2.zip", v2, "application/zip")},
cookies=owner_cookies,
)
assert u.status_code == 422, u.text
assert _hash_tree(plugin_dir) == before_hash, (
"monkey-patched check failure polluted the live tree"
)