No prompt injection or malicious collection endpoint was found; most static matches are Markdown or deployment documentation. The skill still handles secret-bearing environment files, executes mutable package versions, and documents non-interactive production mutations.
bunx @prisma/cli@latest project env add --file .env --role production
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
bunx @prisma/cli@latest project env update --file .env --role production
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
`app deploy --env .env` loads environment variables from a file for the deployment. A config-backed
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
bunx @prisma/cli@latest app deploy api --prod --yes --env .env
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The configuration explicitly loads .env and .env.production files. These variants commonly contain deployment secrets and therefore require deliberate scope and repository controls.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
- an initial Compute deployment with env vars loaded from `.env`
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The command reads a local .env file and prints variable names while redacting values. Redaction reduces exposure, but access to a secret-bearing file is real.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
1. Put a real production-ready `DATABASE_URL` in `.env` or project env.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
4. Redeploy with `--env .env` or project env configured.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
- `env-deploy-loads-dotenv` - Generated deploy scripts may load env via `prisma.compute.ts` or `--en
The guidance configures a local .env file as deployment input, causing secret-bearing values to be read and potentially uploaded to Prisma Compute. This is intended but security-sensitive behavior.
Commands repeatedly execute @prisma/cli@latest and create-prisma@latest through package runners. A compromised or unexpectedly changed release would execute with the agent's local privileges.
The cited commands explicitly use the mutable latest tag with bunx. Package runners download and execute that resolved release.
The skill documents production deploy, promotion, rollback, removal, and other remote mutations with non-interactive confirmation flags. Incorrect target inference could alter live infrastructure.
The cited guidance explicitly combines production operations with --yes. Intent guardrails are present, but the commands bypass interactive confirmation once run.
Capability review items (16)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
The guidance permits a database URL in command arguments, so substituted credentials can enter shell history or process listings. The example is legitimate but security-sensitive.
The guidance permits a database URL in command arguments, so substituted credentials can enter shell history or process listings. The example is legitimate but security-sensitive.
- Existing database env values supplied through `--env DATABASE_URL=...`, `--env DIRECT_URL=...`, an
The guidance permits a database URL in command arguments, so substituted credentials can enter shell history or process listings. The example is legitimate but security-sensitive.
The documented workflow creates, reads, stores, or transmits a real DATABASE_URL. This is intended deployment behavior, but the value is a credential requiring strict redaction and scope controls.
The documented workflow creates, reads, stores, or transmits a real DATABASE_URL. This is intended deployment behavior, but the value is a credential requiring strict redaction and scope controls.
if (!databaseUrl) throw new Error("DATABASE_URL is required")
The documented workflow creates, reads, stores, or transmits a real DATABASE_URL. This is intended deployment behavior, but the value is a credential requiring strict redaction and scope controls.
The documented workflow creates, reads, stores, or transmits a real DATABASE_URL. This is intended deployment behavior, but the value is a credential requiring strict redaction and scope controls.
1. Put a real production-ready `DATABASE_URL` in `.env` or project env.
The documented workflow creates, reads, stores, or transmits a real DATABASE_URL. This is intended deployment behavior, but the value is a credential requiring strict redaction and scope controls.
The configuration or command explicitly references a hidden .env file as deployment input. That file commonly contains credentials and is read during the intended workflow.
The configuration or command explicitly references a hidden .env file as deployment input. That file commonly contains credentials and is read during the intended workflow.
The configuration or command explicitly references a hidden .env file as deployment input. That file commonly contains credentials and is read during the intended workflow.
The configuration or command explicitly references a hidden .env file as deployment input. That file commonly contains credentials and is read during the intended workflow.
The SDK example reads PRISMA_API_TOKEN from the process environment and supplies it to an authenticated client. This is legitimate secret access that requires redaction and least privilege.
The SDK example reads DATABASE_URL from the process environment before deployment. This is legitimate credential handling, but accidental logging or broad process access could expose it.
The SDK example reads PRISMA_API_TOKEN from the process environment and supplies it to an authenticated client. This is legitimate secret access that requires redaction and least privilege.
The SDK example reads DATABASE_URL from the process environment before deployment. This is legitimate credential handling, but accidental logging or broad process access could expose it.
All 226 static findings are false positives caused by documentation examples, Markdown backticks, deployment configuration, and benign scoped checks. No prompt injection, secret exfiltration, or malicious automation was found. The skill does recommend executing unpinned latest-tag packages, which creates a medium supply-chain risk.
The skill recommends running packages through bunx with the latest tag. It also documents an installer that shells out to skills@latest. A compromised or changed upstream release could execute unexpected code.
The cited documentation explicitly uses @latest package execution and describes a secondary latest-tag installer. This is a clear supply-chain exposure, although it is not evidence of a current compromise.
All 226 static findings are false positives caused by documentation examples, Markdown backticks, deployment configuration, and benign scoped checks. No prompt injection, secret exfiltration, or malicious automation was found. The skill does recommend executing unpinned latest-tag packages, which creates a medium supply-chain risk.
The skill recommends running packages through bunx with the latest tag. It also documents an installer that shells out to skills@latest. A compromised or changed upstream release could execute unexpected code.
The cited documentation explicitly uses @latest package execution and describes a secondary latest-tag installer. This is a clear supply-chain exposure, although it is not evidence of a current compromise.
All 226 static findings are false positives caused by documentation examples, Markdown backticks, deployment configuration, and benign scoped checks. No prompt injection, secret exfiltration, or malicious automation was found. The skill does recommend executing unpinned latest-tag packages, which creates a medium supply-chain risk.
The skill recommends running packages through bunx with the latest tag. It also documents an installer that shells out to skills@latest. A compromised or changed upstream release could execute unexpected code.
The cited documentation explicitly uses @latest package execution and describes a secondary latest-tag installer. This is a clear supply-chain exposure, although it is not evidence of a current compromise.
All 226 static findings are false positives caused by documentation examples, Markdown backticks, deployment configuration, and benign scoped checks. No prompt injection, secret exfiltration, or malicious automation was found. The skill does recommend executing unpinned latest-tag packages, which creates a medium supply-chain risk.
The skill recommends running packages through bunx with the latest tag. It also documents an installer that shells out to skills@latest. A compromised or changed upstream release could execute unexpected code.
The cited documentation explicitly uses @latest package execution and describes a secondary latest-tag installer. This is a clear supply-chain exposure, although it is not evidence of a current compromise.
All 226 static findings are false positives caused by documentation examples, Markdown backticks, deployment configuration, and benign scoped checks. No prompt injection, secret exfiltration, or malicious automation was found. The skill does recommend executing unpinned latest-tag packages, which creates a medium supply-chain risk.
The skill recommends running packages through bunx with the latest tag. It also documents an installer that shells out to skills@latest. A compromised or changed upstream release could execute unexpected code.
The cited documentation explicitly uses @latest package execution and describes a secondary latest-tag installer. This is a clear supply-chain exposure, although it is not evidence of a current compromise.
All 226 static findings are false positives caused by documentation examples, Markdown backticks, deployment configuration, and benign scoped checks. No prompt injection, secret exfiltration, or malicious automation was found. The skill does recommend executing unpinned latest-tag packages, which creates a medium supply-chain risk.
The skill recommends running packages through bunx with the latest tag. It also documents an installer that shells out to skills@latest. A compromised or changed upstream release could execute unexpected code.
The cited documentation explicitly uses @latest package execution and describes a secondary latest-tag installer. This is a clear supply-chain exposure, although it is not evidence of a current compromise.