A review of the Taskly codebase against 449 checks across code, secrets, dependencies, infrastructure and release hygiene. Your code was never executed.
VERDICTREBUILD
Client
Taskly (sample)
Prepared by
AIpento, a product of UneeOne
Date
21 September 2026
Report ID
APR-2026-0921-TSK
Method
Static code inspection (SAST)
Classification
Confidential
Confidential — prepared for Taskly — not for distribution without permission
ConfidentialTaskly · synthetic samplePage 1 of 8
Document controlAIpento readiness report
Document control
About this report
Version
1.0
Status
Final
Prepared by
AIpento automated review
Reviewed by
Adjudicator (strict code)
Date of review
21 September 2026
Repository state
main @ sample
Distribution
Taskly security and engineering. Confidential.
Evidence integrity
sha256: 4f2c...e91a (manifest signed; every finding carries its own excerpt hash)
Contents
ConfidentialTaskly · synthetic samplePage 2 of 8
Executive summaryAIpento readiness report
1 · Executive summary
The verdict is REBUILD.
Several critical flaws and no tests to protect a fix. Repairing them one at a time costs more than rebuilding the core on a sound base.
5Critical
5High
5Medium
1Low
2Confirm
What this means for the business. 10 findings need attention before this application carries real customer data. 5 are critical: each one, on its own, lets an outsider reach data or run code they should not. There is no automated test suite, so any fix risks breaking something else unnoticed. We recommend fixing the critical and high findings, adding one test over the path that matters most, and re-running this report before launch.
2 · Scope and method
This report covers the Taskly repository: 11 files, 7 of them code. The review is static: the code was read, not run. No server was started and no request was sent to any system. Findings describe patterns in the code that are known to cause the problems named. A live penetration test, which confirms each finding against a running copy of the app, is a separate engagement.
Checks were drawn from AIpento's catalogue of 449 static checks across 21 lenses: code security (JavaScript, TypeScript, Python, Go, PHP), committed secrets, vulnerable dependencies, infrastructure as code, failure handling, performance and release hygiene. Every check is proven against a broken and a correct example before it may run.
ConfidentialTaskly · synthetic samplePage 3 of 8
Findings by severityAIpento readiness report
3 · Findings by severity
18 findings in all.
Severity is the worst plausible outcome if the finding is left as it is. Two findings are marked needs your confirmation: they depend on intent that no scan can read, and they never decide the verdict on their own.
Critical5
High5
Medium5
Low1
Needs your confirmation2
By lens
Lens
Findings
Security
7
Release
3
Failure
3
Data
2
Speed
2
Tests
1
ConfidentialTaskly · synthetic samplePage 4 of 8
Detailed findingsAIpento readiness report
4 · Detailed findings
Every finding, with its evidence.
Criticalsecurity lensCVSS 9.1
Administrative database key referenced in client-side code
src/lib/supabase.js:6
process.env.SUPABASE_SERVICE_ROLE_KEY,
CWE-522A02:2021 Cryptographic FailuresCheck S7
Why it matters. A service-role key bypasses every row-level rule. In browser code it is visible to every visitor, which makes your whole database readable and writable.
What to do. Use the public key in the browser. Keep the service-role key on the server only, and rotate the exposed one.
Criticalsecurity lensCVSS 9.1
Code built and executed at runtime
src/lib/rules.js:3
return eval(expression);
CWE-95A03:2021 InjectionCheck S4
Why it matters. If any part of the string can be influenced from outside, this runs an attacker code inside your process.
What to do. Replace with a lookup table or a parser. There is almost always a way that does not execute text.
Criticaldata lensCVSS 9.1
Environment file is not excluded from version control
.env:1
CWE-538A05:2021 Security MisconfigurationCheck D2
Why it matters. A .env holds your keys. Unignored, the next commit publishes them to anyone with repository access.
What to do. Add .env to .gitignore, rotate anything it contains, and commit a .env.example with the names only.
Criticalsecurity lensCVSS 9.1
Live credential committed to the codebase
src/lib/ai.js:1
const OPENAI_KEY = 'sk-p[REDACTED]';
CWE-798A07:2021 Identification and Authentication FailuresCheck S1
Why it matters. Anyone who ever gets a copy of this repository has the key. Deleting the line does not help: it stays in git history.
What to do. Rotate the credential first, then move it to an environment variable. Treat it as exposed until it is rotated.
Criticalsecurity lensCVSS 9.1
SQL built by joining strings
api/tasks.js:7
sql: "SELECT * FROM tasks WHERE title LIKE '%" + q + "%'",
CWE-89A03:2021 InjectionCheck S2
Why it matters. A value from a visitor becomes part of the query. This is how a form field turns into "read every row in the table".
What to do. Use parameterised queries. Never concatenate or interpolate a value into SQL.
ConfidentialTaskly · synthetic samplePage 5 of 8
Detailed findingsAIpento readiness report
4 · Detailed findings (continued)
Highsecurity lensCVSS 7.4
Cross-origin requests allowed from anywhere with credentials
api/share.js:2
cors: { origin: '*', credentials: true },
CWE-942A05:2021 Security MisconfigurationCheck S6
Why it matters. Any website can make authenticated calls to your API using your visitor own session.
What to do. Replace the wildcard with an explicit list of origins you control.
Highdata lensCVSS 7.4
Database in use with no migration history
CWE-1059Data integrityCheck D1
Why it matters. The only record of your schema is the live database. A new environment cannot be created, and a lost database cannot be rebuilt.
What to do. Adopt a migration tool and capture the current schema as the first migration.
Why it matters. One hundred items becomes one hundred round trips. This is the single most common reason an AI-built app slows to a crawl after launch.
What to do. Fetch the set in one query and join in memory, or use the database own batch form.
Hightests lensCVSS 7.4
No automated tests anywhere in the codebase
CWE-1120Release hygieneCheck T1
Why it matters. Nothing can tell you whether the next change broke something that used to work. Every fix becomes a gamble, and AI edits are large.
What to do. Start with one test over the path that earns you money. One real test beats a coverage target.
Highsecurity lensCVSS 7.4
Untrusted value written into the page as HTML (2 places)
Why it matters. If that value ever contains a script, it runs in your visitor browser with their session.
What to do. Use textContent, or React children. If HTML is genuinely needed, sanitise it first.
Mediumrelease lensCVSS 5.3
Dependency pinned to a moving target
package.json:6
"@supabase/supabase-js": "*"
CWE-1104A06:2021 Vulnerable and Outdated ComponentsCheck R2
Why it matters. A version of "latest" or "*" means a stranger release can change your app without you touching it.
What to do. Pin to a specific version or a narrow range.
Mediumfailure lensCVSS 5.3
Error caught and discarded
src/lib/sync.js:16
} catch (err) {}
CWE-390A09:2021 Security Logging and Monitoring FailuresCheck F1
Why it matters. The failure still happened; now nobody can see it. These are the bugs that take days to find because there is no trace.
What to do. Log it with enough context to identify the request, and decide deliberately whether to continue or stop.
Mediumfailure lensCVSS 5.3
Network call with no error handling in the file
src/lib/ai.js:4
const res = await fetch('https://api.openai.com/v1/chat/completions', {
CWE-391Failure handlingCheck F2
Why it matters. Other people servers time out and return errors. An unhandled rejection takes down the request, or worse, silently does nothing.
What to do. Wrap the call, decide what the user sees when it fails, and make that path visible.
ConfidentialTaskly · synthetic samplePage 6 of 8
Remediation and coverageAIpento readiness report
5 · Remediation and verification
How a fix is accepted.
When AutoFix proposes a change, it is accepted only when your test suite passes three times in a row, each time from a clean checkout with a fresh database, and no new critical or high finding appears. A fix that passes once is not accepted: the second clean run catches a suite that only passes because of state the first run left behind.
Finding
Severity
Fix available
Verification
SQL built by joining strings
Critical
Parameterised query
Three times, from clean
Value written as HTML
High
Encode on output
Three times, from clean
Live credential committed
Critical
Rotate and move to env
Needs you to rotate the key
Endpoint with no auth
Confirm
Add owner check
Needs you to confirm intent
Fixes that depend on a decision only you can make, such as rotating a key or confirming an endpoint is meant to be public, are never applied automatically.
6 · What was not tested
So this report cannot be mistaken for more than it is:
Nothing was executed. No server was started and no request was sent. Behaviour that only exists at runtime was not seen here.
No live target. Headers, authentication behaviour and rate limits are tested in a separate penetration test against a staging copy you authorise.
Dependencies. npm and PyPI packages were checked against the public OSV database. Packages from other ecosystems were not checked.
Absence is not proof. These checks catch the failures that recur in this kind of software. A determined review finds more.
ConfidentialTaskly · synthetic samplePage 7 of 8
Compliance and glossaryAIpento readiness report
7 · Compliance mapping
Mapped to your controls.
Each finding is mapped to the controls an auditor works from. A control is marked pass only when no open finding maps to it. This is evidence mapped to controls, not a certification. AIpento certifies no one.
Control
Status
Basis
ISO 27001:2022
A.8.28 Secure coding
PARTIAL
Injection and unsafe-eval findings open
A.8.9 Configuration management
FAIL
Secret in code; .env tracked; CORS wildcard
A.8.24 Use of cryptography
FAIL
Admin key reachable from the browser
A.8.16 Monitoring activities
PARTIAL
Errors discarded in one place
SOC 2 (Security)
CC6.1 Logical access
FAIL
Unauthenticated write endpoint; access not checked by owner
CC6.6 Boundary protection
FAIL
CORS allows any origin with credentials
CC7.2 Detection
PARTIAL
A swallowed error hides failures
CC8.1 Change management
PARTIAL
No lockfile; a dependency floats
PCI DSS 4.0
6.2.4 Secure software
FAIL
SQL built from strings; eval of runtime code
6.3.1 Known vulnerabilities
PARTIAL
No lockfile to pin versions
6.4.1 Public-facing apps
FAIL
Reflected HTML without encoding
OWASP ASVS 4.0.3
V5 Validation & encoding
FAIL
Injection and XSS present
V4 Access control
FAIL
Object access not scoped to the owner
V14 Configuration
FAIL
Secret and service key exposed
8 · Glossary
SAST
Static application security testing: reading code without running it.
IDOR / BOLA
Reaching a record by its identifier without a check that it belongs to you.
CWE
A catalogue of software weakness types, each with a number.
CVSS
A standard 0-10 score for how severe a vulnerability is.
A fuller glossary is at aipento.com/glossary. Questions about this report: your AIpento project console.