MCP, MCP server
Engram Memory
Persistent semantic memory for AI agents.
Compatibility: Not stated.
Grade B
Minor issues only, no Medium or High. Worth a quick look at the findings below before installing.
Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.
Rescanned at commit 761a7da48495, unchanged since the check below.
PERMISSION_DELETE_FILES INSTALL_UNPINNED NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV SECRETS_ENV_READ
Why this grade
- PERMISSION_DELETE_FILES Low
Contains a call that can delete files.
plugin.py, line 170
result = subprocess.run(cmd, cwd=str(SCRIPT_DIR), capture_output=True, text=True, timeout=30) - INSTALL_UNPINNED Low
Dependencies use a range or git URL with no lockfile.
requirements.txt, line 1
qdrant-client>=1.7.0 - NETWORK_NO_FIXED_HOST Low
Makes a network call whose host is not written as a fixed address.
SKILL.md, line 338
curl "http://localhost:6333/collections/agent-memory/points/scroll" \ - SECRETS_PATH_DOTENV Low
Mentions a .env file.
.gitignore, line 13
.env - SECRETS_ENV_READ Note
Reads an environment variable whose name contains KEY, SECRET, TOKEN, PASSWORD or CREDENTIAL.
docker/mcp/entrypoint.py, line 56
API_KEY = os.getenv("ENGRAM_API_KEY", "")
Any High, or two or more Medium findings, give D. One Medium gives C. Low findings or a security Note give B. Each code names a finding behind the letter.
Grades are a hygiene check of the source at the commit shown, on the date checked, not a security audit or a guarantee. The free directory grades are automated. Before release, a PlugVet engineer reviews every paid report and any grade we hold back, including all F grades, against the source code and the evidence we quote.
What was checked
Plain notes from the hygiene checks.
Permissions
Concern
can delete files.
PERMISSION_DELETE_FILES
plugin.py, line 170
result = subprocess.run(cmd, cwd=str(SCRIPT_DIR), capture_output=True, text=True, timeout=30)
bridge/project.py, line 52
result = subprocess.run(
scripts/engram_graph.py, line 197
proc = subprocess.run(cmd, capture_output=True, text=True)
Show more
scripts/fastembed_setup.py, line 38
subprocess.run([sys.executable, "-m", "venv", str(self.venv_path)], check=True)
scripts/fastembed_setup.py, line 47
subprocess.run([
scripts/fastembed_setup.py, line 69
result = subprocess.run([
scripts/fastembed_setup.py, line 75
subprocess.run(["docker", "start", "engram-qdrant"], check=True)
scripts/fastembed_setup.py, line 79
subprocess.run([
scripts/fastembed_setup.py, line 226
shutil.rmtree(self.venv_path)
scripts/hooks/engram_hook.py, line 287
os.remove(cache_path)
Install and update
Concern
dependencies are not pinned.
INSTALL_UNPINNED
requirements.txt, line 1
qdrant-client>=1.7.0
Network
Concern
can reach a host that is not written as a fixed address.
NETWORK_NO_FIXED_HOST
SKILL.md, line 338
curl "http://localhost:6333/collections/agent-memory/points/scroll" \
bridge/client.py, line 78
resp = http.get(self._url("/health"))
bridge/client.py, line 142
resp = http.get(
Show more
bridge/client.py, line 209
resp = http.get(
context/tools/semantic_search.py, line 432
print(" 2. Check FastEmbed service: curl http://localhost:11435/health")
docker/all-in-one/init.sh, line 24
if curl -sf http://localhost:6333/healthz >/dev/null 2>&1; then
docker/all-in-one/init.sh, line 32
if curl -sf http://localhost:11435/health >/dev/null 2>&1; then
docker/all-in-one/init.sh, line 41
if ! curl -sf http://localhost:6333/collections/agent-memory >/dev/null 2>&1; then
docker/all-in-one/init.sh, line 42
curl -sf -X PUT http://localhost:6333/collections/agent-memory \
docker/mcp/dashboard.py, line 612
const r = await fetch("/dashboard/timeline?" + params);
docker/mcp/dashboard.py, line 622
const r = await fetch("/search", {method:"POST", headers:{"Content-Type":"application/json"}, body:JSON.stringify({query, limit:20})});
docker/mcp/dashboard.py, line 630
const r = await fetch("/dashboard/stats");
mcp/server.py, line 837
with urllib.request.urlopen(req, timeout=15) as resp:
mcp/server.py, line 943
with urllib.request.urlopen(req, timeout=10) as resp:
scripts/fastembed_setup.py, line 56
response = requests.get("http://localhost:6333/collections", timeout=5)
scripts/setup-context.sh, line 91
if curl -s http://localhost:6333/health > /dev/null 2>&1; then
Secrets
Concern
reads a credential or environment file.
SECRETS_ENV_READ SECRETS_PATH_DOTENV
docker/mcp/entrypoint.py, line 56
API_KEY = os.getenv("ENGRAM_API_KEY", "")
mcp/server.py, line 978
api_key=os.getenv("ENGRAM_API_KEY", ""),
.gitignore, line 13
.env
Show more
plugin/index.js, line 13
const apiKey = cfg.apiKey || process.env.ENGRAM_API_KEY || "";
sdks/python/engrammemory/client.py, line 65
env = os.environ.get("ENGRAM_API_KEY", "")
sdks/typescript/examples/basic-usage.ts, line 15
const apiKey = process.env.ENGRAM_API_KEY;
sdks/typescript/examples/feedback-loop.ts, line 16
const apiKey = process.env.ENGRAM_API_KEY;
sdks/typescript/examples/hive-sharing.ts, line 18
const apiKey = process.env.ENGRAM_API_KEY;
sdks/typescript/examples/hive-sharing.ts, line 35
const keyPrefix = process.env.GRANT_KEY_PREFIX;
.gitignore, line 12
# Environment
.gitignore, line 14
.env.local
README-community.md, line 260
# In your MiroFish .env
context/templates/web_app.yaml, line 566
cp .env.example .env
context/templates/web_app.yaml, line 787
# Check .env file
Agent instructions
Pass
No instruction to ignore earlier instructions or to hide a step.
Obfuscation
Pass
No encoded or minified payload found.
Licence
Pass
A licence file is present (LICENSE).
Maintainer activity
Pass
The latest commit is within the last 365 days.
Note
Hygiene check of commit 761a7da48495. can delete files. dependencies are not pinned. can reach a host that is not written as a fixed address. reads a credential or environment file.
Date and source
Date checked:
github.com
Commit graded: 761a7da48495b0d5ddf7e1d505bf6e8c8f6d7101
This link goes to the original host. PlugVet does not host or check what you download from there, and installing it is your own decision and risk.