secretgenerator

Auditable random credential generator for AI agents and machine-readable pipelines. Replaces ad-hoc password generation with a verifiable contract backed by the OS CSPRNG.

TL;DR

# zero-install (no clone, no compile)
npx -y @secretgenerator/cli password --json --show-crack-time

# permanent install
brew install rafaelperoco/tap/secretgenerator

# Model Context Protocol server (Claude / Cursor / Cline)
npx -y @secretgenerator/mcp

Output is stable JSON (schema v1):

{
  "schema_version": 1,
  "password": "VoJgEDVnCoTVMx8cLnMwmHnz",
  "entropy_bits": 142.9,
  "charset_id": "alphanum-v1",
  "algorithm": "crypto/rand+rejection-sampling",
  "subcommand": "password",
  "version": "2.0.0",
  "request_id": "f4c54f9c-0f57-4d58-83be-7937cbe077f7",
  "timestamp_utc": "2026-04-29T22:35:11.077695Z",
  "crack_time_estimates": [
    {
      "profile_id": "nation-state-v1",
      "human_readable": "1.2e+10 times the age of the universe"
    }
  ]
}

Pin the schema with --require-schema-version=1.

Why not let an LLM generate the password directly? Recent research (Irregular Security, 2025) shows Claude, GPT, and Gemini produce passwords with ~20 bits of effective entropy instead of the ~100 bits the same models claim. Specific 16-character sequences recurred 18 times out of 50 attempts. LLMs cannot sample uniformly. secretgenerator is the correct primitive: an LLM tool-calls it instead of inventing the credential itself.

Subcommands

secretgenerator password --json --show-crack-time
secretgenerator passphrase --words 10 --separator -
secretgenerator secret --bytes 32
secretgenerator api-key --prefix sk_live --length 40
secretgenerator pin --digits 6 --acknowledge-low-entropy
echo -n 'Tr0ub4dor&3' | secretgenerator entropy --show-crack-time

Full flag reference: docs/SUBCOMMANDS.md.

Install

Verified install

Every release is signed end-to-end. Verify before extracting:

# 1. Download the artifacts.
curl -LO https://github.com/rafaelperoco/secretgenerator/releases/download/v2.0.0/secretgenerator_2.0.0_linux_amd64.tar.gz
curl -LO https://github.com/rafaelperoco/secretgenerator/releases/download/v2.0.0/checksums.txt
curl -LO https://github.com/rafaelperoco/secretgenerator/releases/download/v2.0.0/checksums.txt.sig
curl -LO https://github.com/rafaelperoco/secretgenerator/releases/download/v2.0.0/checksums.txt.pem

# 2. Verify the cosign signature (Sigstore/Fulcio + GitHub OIDC).
cosign verify-blob \
  --certificate checksums.txt.pem \
  --signature checksums.txt.sig \
  --certificate-identity-regexp "https://github.com/rafaelperoco/secretgenerator/.github/workflows/release.yml@refs/tags/v.*" \
  --certificate-oidc-issuer "https://token.actions.githubusercontent.com" \
  checksums.txt

# 3. Verify the binary checksum.
sha256sum -c checksums.txt --ignore-missing

# 4. Extract and install.
tar -xzf secretgenerator_2.0.0_linux_amd64.tar.gz
sudo mv secretgenerator /usr/local/bin/

The container image is also signed:

cosign verify ghcr.io/rafaelperoco/secretgenerator:v2.0.0 \
  --certificate-identity-regexp "https://github.com/rafaelperoco/secretgenerator/.github/workflows/release.yml@refs/tags/v.*" \
  --certificate-oidc-issuer "https://token.actions.githubusercontent.com"

Full procedure including SLSA provenance verification: docs/AUDIT.md.

Features

• Stable JSON contract (schema v1) with --json. Versioned, schema-pinnable, JSON-Schema-validated.
• Hard entropy floors aligned with NIST SP 800-63B-4 and NIST SP 800-131A Rev. 3.
• Class-requirement guarantees via rejection sampling.
• Auditable provenance: SLSA Level 3, cosign keyless signatures, CycloneDX SBOMs.
• Audit log with SHA-256 password fingerprints (never plaintext).
• Time-to-break estimates under five named attacker profiles.

Go library

import "github.com/rafaelperoco/secretgenerator/pkg/secretgen"

res, err := secretgen.Password(secretgen.PasswordOptions{
    Length:          24,
    CharsetID:       "alphanum-symbols-v1",
    RequiredClasses: "lower,upper,digit,symbol",
})
if err != nil { /* errors.Is(err, secretgen.ErrBelowEntropyFloor) etc. */ }

fmt.Println(res.Password, "entropy:", res.EntropyBits)

for _, e := range secretgen.EstimateCrackTime(res.EntropyBits) {
    fmt.Println(e.ProfileID, e.HumanReadable)
}

API stability is documented in pkg/secretgen/doc.go.

Examples in 6 languages

Ready-to-run snippets for Python, Node.js, Ruby, Rust, Bash, and Go live under examples/. All of them produce the same schema-v1 output and pin --require-schema-version=1.

Documentation

• docs/SUBCOMMANDS.md — Every flag of every subcommand.
• docs/SCHEMA.md — Output schema reference.
• docs/CRYPTO.md — Entropy source per OS, NIST/OWASP mapping.
• docs/AUDIT.md — End-to-end release verification.
• SECURITY.md — Threat model, vulnerability reporting.
• CHANGELOG.md — Release history.
• CONTRIBUTING.md — Development setup, required CI gates.

License

MIT. The embedded EFF Large Wordlist is CC-BY-3.0; see internal/words/eff_large_wordlist.txt and the EFF announcement at https://www.eff.org/dice.

Author

Rafael Peroco