Skills adaptyv Audit History
๐Ÿ“ฆ

Audit History

adaptyv - 9 audits

Version comparison

Capability and finding changes across audited versions, newest first.

VersionDateResultReview itemsChange vs previous
v9 LatestJul 9, 2026, 04:11 AM 1 confirmed20No capability change
v8 Jul 9, 2026, 04:11 AM 1 confirmed20No capability change
v7 Jul 5, 2026, 05:58 PM 1 confirmed22No capability change
v6 Jul 5, 2026, 05:58 PM 1 confirmed22No capability change
v5 Jun 30, 2026, 05:31 AM 3 confirmed1Filesystem accessExternal commands Contains scripts
v4 Jan 17, 2026, 06:49 AM No confirmed findings0No capability change
v3 Jan 17, 2026, 06:49 AM No confirmed findings0Contains scripts External commandsFilesystem access
v2 Jan 12, 2026, 04:04 PM No confirmed findings0External commands
v1 Jan 4, 2026, 04:33 PM No confirmed findings0Baseline

Jul 9, 2026, 04:11 AM

Most high-severity static matches are false positives caused by placeholder API keys, .env guidance, Markdown fences, or scientific sequence terminology. The real risks are intended outbound API calls to Adaptyv and NetSolP, local example file writes, and disclosure of proprietary protein sequences to a third-party lab service.

5
Files scanned
2,336
Lines analyzed
25
Review items
0
False positives ignored

Confirmed security concerns (1)

Medium
Third-party submission of proprietary protein sequences
The skill is designed to submit protein sequences and experiment metadata to the Adaptyv cloud lab API. This can expose proprietary biological designs or research plans if users do not have permission to share them with that service.
The documentation directly describes submitting protein sequences through the Adaptyv API and shows request payloads containing sequences and webhooks. The risk is contextual data disclosure, not hidden malicious behavior.
Capability review items (20)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Medium
Python file write/append
with open(output_file, 'w') as f:
The example writes experiment result data to local files using API-derived identifiers or caller-provided output paths. This is normal example behavior, but it can overwrite local files if copied without path controls.
Medium
Python file write/append
with open(f"{experiment_id}_info.json", 'w') as f:
The example writes experiment result data to local files using API-derived identifiers or caller-provided output paths. This is normal example behavior, but it can overwrite local files if copied without path controls.
Medium
Python os file operations
os.makedirs(output_dir, exist_ok=True)
The example writes experiment result data to local files using API-derived identifiers or caller-provided output paths. This is normal example behavior, but it can overwrite local files if copied without path controls.
Low
Hardcoded URL
https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Python HTTP libraries
response = requests.get(f"{BASE_URL}/organization/credits", headers=HEADERS)
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Python HTTP libraries ยท 6 occurrences
response = requests.post(
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Python HTTP libraries ยท 4 occurrences
response = requests.get(
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Hardcoded URL ยท 2 occurrences
BASE_URL = "https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws"
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Python HTTP libraries
response = requests.post(url, data=data)
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Hardcoded URL
url = "https://services.healthtech.dtu.dk/services/NetSolP-1.0/api/predict"
The example sends protein sequence data to an external DTU NetSolP web service. This is intended scientific functionality, but it is still a real outbound data-sharing action.
Low
Hardcoded URL
"https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws/experiments",
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.

Risk Factors

๐ŸŒ Network access (28)
๐Ÿ“ Filesystem access (4)
๐Ÿ”‘ Env variables (24)
โš™๏ธ External commands (12)
Audited by: claude

Jul 9, 2026, 04:11 AM

Most high-severity static matches are false positives caused by placeholder API keys, .env guidance, Markdown fences, or scientific sequence terminology. The real risks are intended outbound API calls to Adaptyv and NetSolP, local example file writes, and disclosure of proprietary protein sequences to a third-party lab service.

5
Files scanned
2,336
Lines analyzed
25
Review items
0
False positives ignored

Confirmed security concerns (1)

Medium
Third-party submission of proprietary protein sequences
The skill is designed to submit protein sequences and experiment metadata to the Adaptyv cloud lab API. This can expose proprietary biological designs or research plans if users do not have permission to share them with that service.
The documentation directly describes submitting protein sequences through the Adaptyv API and shows request payloads containing sequences and webhooks. The risk is contextual data disclosure, not hidden malicious behavior.
Capability review items (20)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Medium
Python file write/append
with open(output_file, 'w') as f:
The example writes experiment result data to local files using API-derived identifiers or caller-provided output paths. This is normal example behavior, but it can overwrite local files if copied without path controls.
Medium
Python file write/append
with open(f"{experiment_id}_info.json", 'w') as f:
The example writes experiment result data to local files using API-derived identifiers or caller-provided output paths. This is normal example behavior, but it can overwrite local files if copied without path controls.
Medium
Python os file operations
os.makedirs(output_dir, exist_ok=True)
The example writes experiment result data to local files using API-derived identifiers or caller-provided output paths. This is normal example behavior, but it can overwrite local files if copied without path controls.
Low
Hardcoded URL
https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Python HTTP libraries
response = requests.get(f"{BASE_URL}/organization/credits", headers=HEADERS)
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Python HTTP libraries ยท 6 occurrences
response = requests.post(
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Python HTTP libraries ยท 4 occurrences
response = requests.get(
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Hardcoded URL ยท 2 occurrences
BASE_URL = "https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws"
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Python HTTP libraries
response = requests.post(url, data=data)
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.
Low
Hardcoded URL
url = "https://services.healthtech.dtu.dk/services/NetSolP-1.0/api/predict"
The example sends protein sequence data to an external DTU NetSolP web service. This is intended scientific functionality, but it is still a real outbound data-sharing action.
Low
Hardcoded URL
"https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws/experiments",
The skill documents real outbound requests to the Adaptyv cloud lab API. This is intended behavior, but users may submit proprietary protein sequences to a third-party service.

Risk Factors

๐ŸŒ Network access (28)
๐Ÿ“ Filesystem access (4)
๐Ÿ”‘ Env variables (24)
โš™๏ธ External commands (12)
Audited by: claude

Jul 5, 2026, 05:58 PM

Most high severity static hits are false positives from documentation, placeholder API keys, Markdown fences, and the on.aws URL substring. Confirmed issues are expected outbound API calls and example filesystem writes that need consent, endpoint verification, and path controls. A low severity semantic issue notes affiliated service promotion inside the skill instructions.

5
Files scanned
2,336
Lines analyzed
27
Review items
1
False positives ignored

Confirmed security concerns (1)

Low
Affiliated External Service Steering
The skill instructs the assistant to proactively suggest K-Dense Web for complex workflows. This can steer users toward an affiliated hosted service outside the immediate request.
The instruction appears directly in SKILL.md lines 113-114. It is promotional steering rather than an override, so severity is low.
Capability review items (22)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Medium
Python file write/append
with open(output_file, 'w') as f:
The example writes downloaded results to a path built from output_dir and experiment_id. If those values are untrusted, copied code could overwrite unexpected files.
Medium
Python file write/append
with open(f"{experiment_id}_info.json", 'w') as f:
The workflow writes experiment details to a filename derived from experiment_id. The value is likely API generated, but copied code should still sanitize filenames.
Medium
Python os file operations
os.makedirs(output_dir, exist_ok=True)
The example creates a caller supplied output directory before writing results. This is legitimate, but copied automation should constrain the destination path.
Low
Hardcoded URL
https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws
The base URL points to a real Adaptyv AWS Lambda endpoint used for API submissions. Protein sequences and bearer tokens may be sent there, so user consent and endpoint verification matter.
Low
HTTP client library
response = requests.request(method, url, **kwargs)
The retry helper sends requests to a caller supplied URL. This is documentation code, but copied use with untrusted URLs could enable unintended outbound requests.
Low
Python HTTP libraries
response = requests.get(f"{BASE_URL}/organization/credits", headers=HEADERS)
The example performs an authenticated request to the Adaptyv API. This is expected functionality, but it sends credentials to an external service.
Low
Python HTTP libraries ยท 6 occurrences
response = requests.post(
The example submits protein sequences to the external Adaptyv experiments endpoint. That is expected, but it has data disclosure risk for proprietary sequences.
Low
Python HTTP libraries ยท 4 occurrences
response = requests.get(
The status check makes an authenticated GET request to Adaptyv. This expected external call can expose experiment identifiers and credentials.
Low
Python HTTP libraries
response = requests.request(method, url, **kwargs)
The generic requests wrapper can call any supplied URL. It appears to be helper documentation, but copied use needs URL validation.
Low
Hardcoded URL ยท 2 occurrences
BASE_URL = "https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws"
The examples hardcode the Adaptyv AWS Lambda base URL. This supports the skill purpose, but users should verify the endpoint before sending credentials or sequences.
Low
Python HTTP libraries
response = requests.post(url, data=data)
The NetSolP example posts protein sequences to an external DTU web service. This can disclose proprietary sequences outside Adaptyv.
Low
Hardcoded URL
url = "https://services.healthtech.dtu.dk/services/NetSolP-1.0/api/predict"
The URL targets the DTU NetSolP prediction service. Users should approve this third-party sequence transfer before running the example.
Low
Hardcoded URL
"https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws/experiments",
The URL targets the Adaptyv experiment submission endpoint. This is intended API use, but it should be verified before sending credentials or sequences.
Static false positives ignored (1)

These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.

Critical
AWS credentials directory
"https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws/experiments",
Force-confirmed sensitive/critical static finding; AI dismissal overridden.

Risk Factors

๐ŸŒ Network access (28)
๐Ÿ“ Filesystem access (4)
๐Ÿ”‘ Env variables (24)
โš™๏ธ External commands (12)
Audited by: codex

Jul 5, 2026, 05:58 PM

Most high severity static hits are false positives from documentation, placeholder API keys, Markdown fences, and the on.aws URL substring. Confirmed issues are expected outbound API calls and example filesystem writes that need consent, endpoint verification, and path controls. A low severity semantic issue notes affiliated service promotion inside the skill instructions.

5
Files scanned
2,336
Lines analyzed
27
Review items
1
False positives ignored

Confirmed security concerns (1)

Low
Affiliated External Service Steering
The skill instructs the assistant to proactively suggest K-Dense Web for complex workflows. This can steer users toward an affiliated hosted service outside the immediate request.
The instruction appears directly in SKILL.md lines 113-114. It is promotional steering rather than an override, so severity is low.
Capability review items (22)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

Medium
Python file write/append
with open(output_file, 'w') as f:
The example writes downloaded results to a path built from output_dir and experiment_id. If those values are untrusted, copied code could overwrite unexpected files.
Medium
Python file write/append
with open(f"{experiment_id}_info.json", 'w') as f:
The workflow writes experiment details to a filename derived from experiment_id. The value is likely API generated, but copied code should still sanitize filenames.
Medium
Python os file operations
os.makedirs(output_dir, exist_ok=True)
The example creates a caller supplied output directory before writing results. This is legitimate, but copied automation should constrain the destination path.
Low
Hardcoded URL
https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws
The base URL points to a real Adaptyv AWS Lambda endpoint used for API submissions. Protein sequences and bearer tokens may be sent there, so user consent and endpoint verification matter.
Low
HTTP client library
response = requests.request(method, url, **kwargs)
The retry helper sends requests to a caller supplied URL. This is documentation code, but copied use with untrusted URLs could enable unintended outbound requests.
Low
Python HTTP libraries
response = requests.get(f"{BASE_URL}/organization/credits", headers=HEADERS)
The example performs an authenticated request to the Adaptyv API. This is expected functionality, but it sends credentials to an external service.
Low
Python HTTP libraries ยท 6 occurrences
response = requests.post(
The example submits protein sequences to the external Adaptyv experiments endpoint. That is expected, but it has data disclosure risk for proprietary sequences.
Low
Python HTTP libraries ยท 4 occurrences
response = requests.get(
The status check makes an authenticated GET request to Adaptyv. This expected external call can expose experiment identifiers and credentials.
Low
Python HTTP libraries
response = requests.request(method, url, **kwargs)
The generic requests wrapper can call any supplied URL. It appears to be helper documentation, but copied use needs URL validation.
Low
Hardcoded URL ยท 2 occurrences
BASE_URL = "https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws"
The examples hardcode the Adaptyv AWS Lambda base URL. This supports the skill purpose, but users should verify the endpoint before sending credentials or sequences.
Low
Python HTTP libraries
response = requests.post(url, data=data)
The NetSolP example posts protein sequences to an external DTU web service. This can disclose proprietary sequences outside Adaptyv.
Low
Hardcoded URL
url = "https://services.healthtech.dtu.dk/services/NetSolP-1.0/api/predict"
The URL targets the DTU NetSolP prediction service. Users should approve this third-party sequence transfer before running the example.
Low
Hardcoded URL
"https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws/experiments",
The URL targets the Adaptyv experiment submission endpoint. This is intended API use, but it should be verified before sending credentials or sequences.
Static false positives ignored (1)

These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.

Critical
AWS credentials directory
"https://kq5jp7qj7wdqklhsxmovkzn4l40obksv.lambda-url.eu-central-1.on.aws/experiments",
Force-confirmed sensitive/critical static finding; AI dismissal overridden.

Risk Factors

๐ŸŒ Network access (28)
๐Ÿ“ Filesystem access (4)
๐Ÿ”‘ Env variables (24)
โš™๏ธ External commands (12)
Audited by: codex

Jun 30, 2026, 05:31 AM

The scanner reported many command-execution and blocker findings, but review shows most are Markdown code fences, inline endpoint notation, protein-science terms, or example API documentation rather than executable skill code. Confirmed risks are legitimate but meaningful: the skill asks agents to use bearer-token authentication, call external web services, optionally configure webhooks, and write downloaded experiment data to local files. No prompt injection attempt or confirmed malicious intent was found.

5
Files scanned
2,336
Lines analyzed
8
Review items
4
False positives ignored

Confirmed security concerns (3)

Medium
Authenticated External API Access
The skill instructs users and agents to send protein sequences and metadata to the Adaptyv cloud API using bearer-token authentication. This is core functionality, but it can expose proprietary sequences, target choices, webhook URLs, and organization account data if used without user approval.
The network calls and authorization headers are explicit and semantically match the skill purpose. The risk is data exposure through legitimate external API use, not hidden exfiltration.
Medium
Environment-Based Secret Handling
The documentation directs users to store ADAPTYV_API_KEY in an environment variable or .env file and then load it into request headers. This is a normal API pattern, but agents must avoid printing, committing, or transmitting the token outside the intended Adaptyv requests.
The files clearly show API-key setup, dotenv loading, and bearer-header construction. This confirms credential access, but the instructions include basic safe-storage guidance.
Medium
Third-Party Bioinformatics Service Calls
The optimization guidance includes requests to external prediction services before Adaptyv submission. These calls may disclose protein sequences to third-party services and should be treated as user-approved network activity.
The NetSolP request example explicitly sends sequence data to a public service. This is legitimate scientific workflow guidance but has clear data-sharing implications.
Needs review findings (1)

These findings came from uncertain legacy audit verdicts, so they require review but are not counted as confirmed security issues.

Medium
Local Result File Writes Use Dynamic Names
Example workflows create directories and write JSON result files using experiment identifiers. This is useful for analysis, but implementations should sanitize experiment IDs and constrain output directories before writing files.
The file writes are visible in examples and appear intended for local result storage. The risk depends on whether experiment IDs or output paths are trusted in a real implementation.
Static false positives ignored (4)

These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.

Low
Markdown Formatting Misclassified as Ruby Backticks
Most external command findings are false positives caused by Markdown fenced code blocks and inline code spans in documentation. No Ruby backtick execution or active shell execution code was found in the reviewed skill files.
The cited lines are documentation snippets and Markdown formatting. They do not execute as Ruby or shell code inside the skill itself.
Low
Weak Cryptography Flags Are Domain-Term False Positives
The weak cryptography detections appear to match scientific notation and protein-assay terms such as KD, ESM, and ipTM rather than cryptographic algorithms. No evidence found of MD5, SHA1, or other weak cryptographic operations in the inspected context.
The reviewed line ranges describe protein binding metrics and model names, not cryptographic hashing. The static category is therefore not semantically supported.
Low
AWS Credentials Directory Flag Is URL Text
The critical AWS credentials directory finding is a false positive against the Adaptyv Lambda URL containing on.aws. The line is an HTTPS API endpoint, not access to a local AWS credentials directory.
The cited location is a hardcoded HTTPS endpoint and request header example. It does not reference ~/.aws or credential files.
Low
No Prompt Injection Attempt Found
Search and review did not find text that attempts to override evaluator instructions, claim pre-approval, or impersonate system messages. No evidence found for prompt injection in the inspected files.
A targeted review for common injection phrases returned no matches. The final section promotes K-Dense Web but does not instruct the evaluator to skip analysis or change policy.

Detected Patterns

Bearer Token Sent in HTTP Authorization HeaderExternal Submission of Protein Sequence DataDynamic Local JSON Output Paths
Audited by: codex

Jan 17, 2026, 06:49 AM

All 304 static findings are FALSE POSITIVES. The skill is legitimate scientific software for protein research. Scanner patterns triggered on: FASTA format headers (protein sequences in markdown code blocks), API calls to official platform endpoints, environment variable usage for API keys (recommended practice), and scientific notation (e.g., binding affinity values like 1.2e-9). No malicious behavior confirmed.

6
Files scanned
3,608
Lines analyzed
3
Review items
0
False positives ignored
Audited by: claude

Jan 17, 2026, 06:49 AM

All 304 static findings are FALSE POSITIVES. The skill is legitimate scientific software for protein research. Scanner patterns triggered on: FASTA format headers (protein sequences in markdown code blocks), API calls to official platform endpoints, environment variable usage for API keys (recommended practice), and scientific notation (e.g., binding affinity values like 1.2e-9). No malicious behavior confirmed.

6
Files scanned
3,608
Lines analyzed
3
Review items
0
False positives ignored
Audited by: claude

Jan 12, 2026, 04:04 PM

All 287 static findings are false positives. The scanner detected markdown code block delimiters as shell commands, URL mentions as network vulnerabilities, and API key documentation as credential exposure. This is legitimate protein design documentation with example code snippets for a scientific cloud laboratory platform.

5
Files scanned
2,336
Lines analyzed
4
Review items
0
False positives ignored

Risk Factors

โš™๏ธ External commands (144)
reference/api_reference.md:5-7 reference/api_reference.md:7-13 reference/api_reference.md:13-15 reference/api_reference.md:15-23 reference/api_reference.md:23-25 reference/api_reference.md:25 reference/api_reference.md:25-35 reference/api_reference.md:35-38 reference/api_reference.md:38-49 reference/api_reference.md:49-57 reference/api_reference.md:57-64 reference/api_reference.md:64-70 reference/api_reference.md:70-73 reference/api_reference.md:73-84 reference/api_reference.md:84-87 reference/api_reference.md:87-88 reference/api_reference.md:88-89 reference/api_reference.md:89-90 reference/api_reference.md:90-96 reference/api_reference.md:96-99 reference/api_reference.md:99-100 reference/api_reference.md:100-101 reference/api_reference.md:101-104 reference/api_reference.md:104-118 reference/api_reference.md:118-126 reference/api_reference.md:126-129 reference/api_reference.md:129-152 reference/api_reference.md:152-160 reference/api_reference.md:160-163 reference/api_reference.md:163-164 reference/api_reference.md:164-165 reference/api_reference.md:165-168 reference/api_reference.md:168-181 reference/api_reference.md:181-187 reference/api_reference.md:187-190 reference/api_reference.md:190-197 reference/api_reference.md:197-205 reference/api_reference.md:205-208 reference/api_reference.md:208-215 reference/api_reference.md:215-222 reference/api_reference.md:222-230 reference/api_reference.md:230-233 reference/api_reference.md:233-234 reference/api_reference.md:234-235 reference/api_reference.md:235-236 reference/api_reference.md:236-246 reference/api_reference.md:246-258 reference/api_reference.md:258-261 reference/api_reference.md:261-262 reference/api_reference.md:262-263 reference/api_reference.md:263-264 reference/api_reference.md:264-265 reference/api_reference.md:265-266 reference/api_reference.md:266-267 reference/api_reference.md:267-276 reference/api_reference.md:276-279 reference/api_reference.md:279-299 reference/api_reference.md:299-301 reference/examples.md:7-36 reference/examples.md:36-40 reference/examples.md:40-41 reference/examples.md:41-43 reference/examples.md:43-46 reference/examples.md:46-48 reference/examples.md:48-54 reference/examples.md:54-98 reference/examples.md:98-102 reference/examples.md:102-160 reference/examples.md:160-164 reference/examples.md:164-203 reference/examples.md:203-209 reference/examples.md:209-242 reference/examples.md:242-246 reference/examples.md:246-280 reference/examples.md:280-284 reference/examples.md:284-316 reference/examples.md:316-322 reference/examples.md:322-365 reference/examples.md:365-369 reference/examples.md:369-413 reference/examples.md:413-417 reference/examples.md:417-453 reference/examples.md:453-459 reference/examples.md:459-499 reference/examples.md:499-503 reference/examples.md:503-550 reference/examples.md:550-556 reference/examples.md:556-623 reference/examples.md:623-627 reference/examples.md:627-692 reference/examples.md:692-696 reference/examples.md:696-747 reference/examples.md:747-753 reference/examples.md:753-820 reference/examples.md:820-826 reference/examples.md:826-875 reference/examples.md:875-879 reference/examples.md:879-913 reference/experiments.md:66-86 reference/experiments.md:86-128 reference/experiments.md:128-142 reference/experiments.md:142-187 reference/experiments.md:187-201 reference/experiments.md:201-252 reference/experiments.md:252-271 reference/protein_optimization.md:22-31 reference/protein_optimization.md:31-71 reference/protein_optimization.md:71-91 reference/protein_optimization.md:91-110 reference/protein_optimization.md:110-135 reference/protein_optimization.md:135-154 reference/protein_optimization.md:154-188 reference/protein_optimization.md:188-213 reference/protein_optimization.md:213-249 reference/protein_optimization.md:249-263 reference/protein_optimization.md:263-279 reference/protein_optimization.md:279-288 reference/protein_optimization.md:288-327 reference/protein_optimization.md:327-347 reference/protein_optimization.md:347-397 reference/protein_optimization.md:397-414 reference/protein_optimization.md:414-427 reference/protein_optimization.md:427-431 reference/protein_optimization.md:431-454 reference/protein_optimization.md:454-458 reference/protein_optimization.md:458-486 reference/protein_optimization.md:486-490 reference/protein_optimization.md:490-510 reference/protein_optimization.md:510-514 reference/protein_optimization.md:514-571 reference/protein_optimization.md:571-588 reference/protein_optimization.md:588-611 SKILL.md:23-25 SKILL.md:25-27 SKILL.md:27-29 SKILL.md:29-31 SKILL.md:31-37 SKILL.md:37-39 SKILL.md:39-45 SKILL.md:45-72 SKILL.md:72-81 SKILL.md:81-98 SKILL.md:98-101 SKILL.md:101-104
๐ŸŒ Network access (29)
๐Ÿ“ Filesystem access (4)
๐Ÿ”‘ Env variables (31)
Audited by: claude

Jan 4, 2026, 04:33 PM

Documentation-only skill providing guidance for using the Adaptyv API. Contains code examples that make HTTP requests to documented endpoints using API keys from environment variables. No executable scripts, obfuscation, persistence mechanisms, or hidden code paths found.

8
Files scanned
2,629
Lines analyzed
3
Review items
0
False positives ignored

Risk Factors

๐ŸŒ Network access (2)
๐Ÿ”‘ Env variables (2)
๐Ÿ“ Filesystem access (1)
Audited by: claude