CVE-2023-47128 is a critical-severity (CVSS 9.1) SQL Injection vulnerability in piccolo. 1 public exploit reference exists, so weaponization risk is real. A fix is available for piccolo — see the affected versions and patch details below.
piccolo SQL Injection via named transaction savepoints
Exploitation Status
Proof-of-concept exploit code exists
- CISA’s SSVC triage found public proof-of-concept exploit code for this CVE, though no confirmed active exploitation.
- CISA assesses this as automatable — exploitation doesn’t require manual, per-target effort, which raises the odds of mass scanning and opportunistic attacks.
- A successful exploit gives an attacker total control of the affected component, not partial access.
Exploitation and automatability from CISA’s SSVC triage for CVE-2023-47128.
EPSS Exploitation Probability
EPSS (Exploit Prediction Scoring System) is a daily probability model maintained by FIRST.org. It estimates the likelihood a CVE will be exploited in production environments within the next 30 days, derived from real-world threat intelligence signals.
How urgent is this, really
CVE-2023-47128 plotted by exploitation likelihood (EPSS) against impact (CVSS). The shaded corner — EPSS 50%+ and CVSS 7.0+ — is where this CVE doesn't sit, though severity or exploitability alone can still warrant action.
Where this sits among everything scored
Of 378,567 CVEs with a current EPSS score, this one falls in the < 10% band (highlighted). Real counts from FIRST.org, not a sample — log-scaled since the landscape is heavily right-skewed.
Real-World Exposure
piccoloReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects PyPI packages — download data is not available via public APIs for these ecosystems.
Description
Summary
The handling of named transaction savepoints in all database implementations is vulnerable to SQL Injection as user provided input is passed directly to connection.execute(...) via f-strings.
Details
An excerpt of the Postgres savepoint handling:
async def savepoint(self, name: t.Optional[str] = None) -> Savepoint:
name = name or f"savepoint_{self.get_savepoint_id()}"
await self.connection.execute(f"SAVEPOINT {name}")
return Savepoint(name=name, transaction=self)
In this example, we can see user input is directly passed to connection.execute without being properly escaped.
All implementations of savepoints and savepoint methods directly pass this name parameter to connection.execute and are vulnerable to this. A non-exhaustive list can be found below:
Care should be given to ensuring all strings passed to connection.execute are properly escaped, regardless of how end user facing they may be.
Further to this, the following method also passes user input directly to an execution context however I have been unable to abuse this functionality at the time of writing. This method also has a far lower chance of being exposed to an end user as it relates to database init functionality.
PoC
The following FastAPI route can be used in conjunction with sqlmap to easily demonstrate the SQL injection.
DB = ...
@app.get("/test")
async def test(name):
async with DB.transaction() as transaction:
await transaction.savepoint(name)
Steps
- Create a standard Piccolo application with Postgres as a database backend
- Add the route shown previously
- Run your application, making a note of the URL it is served on
- Install sqlmap
- In a terminal, run the following command substituting URL with your applications URL:
sqlmap -u "http://URL/test?name=a" --batch - Observe sqlmap identifying the vulnerability
For sqlmap help, this usage guide may be useful. The following commands may also be helpful to see the impact.
Dumping all tables
The --tables flag will enumerate all tables accessible from within the exposed database session.
sqlmap -u "http://URL/test?name=a" --batch --tables
An example output of this can be seen in the following screenshot.

OS Shell
The --os-shell will drop the user into an OS shell on the underlying system if permissions permit. This can be seen in the attached screenshot which prints the databases current working directory.

Impact
While the likelihood of an end developer exposing a savepoints name parameter to a user is highly unlikely, it would not be unheard of. If a malicious user was able to abuse this functionality they would have essentially direct access to the database and the ability to modify data to the level of permissions associated with the database user.
A non exhaustive list of actions possible based on database permissions is:
- Read all data stored in the database, including usernames and password hashes
- Insert arbitrary data into the database, including modifying existing records
- Gain a shell on the underlying server
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐍PyPI | piccolo | all versions | 1.1.1pip install --upgrade 'piccolo==1.1.1' |
Research use only. For defensive security, authorized penetration testing, and academic research only. Never execute exploit code against systems without explicit written authorization.
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for piccolo, including transitive dependencies — a direct dependency you never call can still pull in a vulnerable version.
Fix
Update piccolo to 1.1.1 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms CVE-2023-47128 is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 Security's impact-aware SCA analyses which vulnerable code paths your application actually calls, so a match like CVE-2023-47128 can be triaged on real exposure rather than presence alone.
Tailored to CVE-2023-47128. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is CVE-2023-47128 in your dependencies?
O3 Security finds CVE-2023-47128 across PyPI dependencies, including transitive ones, and its impact-aware SCA ranks findings by whether your code actually calls the vulnerable path.