Skip to main content

Semantic security scanner for Algorand smart contracts and workflows

Project description

AlgoSec (A Secure)

AlgoSec is a semantic security scanner and runtime decision assistant for Algorand smart contracts.

It supports static contract analysis, on-chain checks, AI-assisted risk narratives, contract auto-remediation, and API-based policy gating.

1) What AlgoSec does today

Contract scanning (TEAL + PyTeal)

  • Scans files or folders containing .teal / .py contracts.
  • Normalizes PyTeal patterns (approval_program / clear_state_program) before analysis.
  • Runs semantic checks for common exploit paths:
    • missing RekeyTo == ZeroAddress
    • missing CloseRemainderTo == ZeroAddress
    • missing AssetCloseTo == ZeroAddress
    • unguarded update/delete/admin paths
    • unsafe Gtxn assumptions without strict group validation
    • missing argument-bound checks
    • state mutation without sender/role authorization

Scoring and risk output

  • Computes security_score, risk_score, grade, and deploy verdict.
  • Exit code is non-zero when findings/errors exist (CI-friendly behavior).

Reports

  • Console dashboard (visual)
  • JSON
  • Markdown
  • SARIF
  • HTML (expanded analytics + charts)

On-chain and indexer analysis

  • Scan deployed app by app_id
  • Scan account risk posture
  • Scan ASA metadata/configuration
  • Scan transaction streams for anomalies

AI-assisted workflow

  • ai-run: scan + simulate + policy decision + explanation artifacts.
  • Works in two modes:
    • deterministic mode (always available; no key required)
    • Gemini narrative mode (optional; enabled when API key is available)

Auto-fix workflow

  • fix-contract applies hardened guard patterns to vulnerable PyTeal contracts.
  • Optional post-fix verification scan (--verify).

Entitlement / premium support

  • premium-status supports ASA-balance mode and subscription-app mode.
  • Testnet deploy commands are dry-run by default for safety.

2) Installation

Standard install

pip install algo-secure

With Gemini SDKs

pip install "algo-secure[ai]"

Development install

pip install -e .

3) Quick start

# Home UI
algosec

# Core scan
algosec analyze ./contracts

# Export html report
algosec scan ./contracts -f html -o report.html

# Compare two versions
algosec diff ./contracts_old ./contracts_new

# Auto-fix a vulnerable contract
algosec fix-contract examples/incoming_contracts/random_user_contract.py --verify

Aliases: algosec, alg, aplussec.

4) Command groups

Static and project scan

  • scan, alg_scan, analyze
  • stats
  • diff

On-chain/indexer scan

  • scan-deployed-app, alg_scan_deployed_app
  • scan-account, alg_scan_account
  • scan-asset, alg_scan_asset
  • scan-transactions, alg_scan_transactions

Testing

  • test-pyteal, alg_test_pyteal

AI and remediation

  • ai-status
  • ai-run, alg_ai_run
  • fix-contract, alg_fix_contract

Entitlement and deploy

  • premium-status
  • deploy-token-registry-testnet
  • deploy-subscription-testnet

API server

  • serve

5) AI behavior (important)

AlgoSec is designed to work on any system even without LLM credentials.

  • If no Gemini key/sdk is available, ai-run still executes using deterministic reasoning (narrative_provider=deterministic).
  • If Gemini is available, narrative generation switches to narrative_provider=gemini.

Gemini key lookup order:

  1. ALGOSEC_GEMINI_API_KEY
  2. GEMINI_API_KEY
  3. GOOGLE_API_KEY

Optional one-shot override:

algosec ai-status --gemini-api-key "<key>"

6) Security and secret handling

  • No API key is hardcoded in source.
  • Auth signing secret does not default to a fixed insecure string.
  • If ALGOSEC_AUTH_SECRET is absent, runtime uses an ephemeral in-memory secret.
  • Recommendation for production API deployments:
    • set ALGOSEC_AUTH_SECRET
    • set ALGOSEC_REQUIRE_API_KEY=1
    • set ALGOSEC_API_KEY

7) API surface

Primary routes include:

  • /health
  • /scan
  • /simulate
  • /safe_execute
  • /agent/run
  • /auth/challenge
  • /auth/verify
  • /auth/me

8) Packaging and release status

  • pyproject.toml is configured for setuptools build.
  • Wheel + sdist build is validated (python -m build, twine check).
  • PyPI publishing workflow is present in .github/workflows/release.yml.

9) License

MIT — see LICENSE.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

algo_secure-0.1.4.tar.gz (74.6 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

algo_secure-0.1.4-py3-none-any.whl (103.6 kB view details)

Uploaded Python 3

File details

Details for the file algo_secure-0.1.4.tar.gz.

File metadata

  • Download URL: algo_secure-0.1.4.tar.gz
  • Upload date:
  • Size: 74.6 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.2

File hashes

Hashes for algo_secure-0.1.4.tar.gz
Algorithm Hash digest
SHA256 38b3e98f2bb9c9d6fc79319b435c9cd65cdc478135ff04f0b04d66b67431eae9
MD5 1164dd72f680bafd57e5e337bd04c46b
BLAKE2b-256 09d94721bae63ab6a16e894739c06cb2e3ea84fe2f88b5086c6cbf5a18e1e067

See more details on using hashes here.

File details

Details for the file algo_secure-0.1.4-py3-none-any.whl.

File metadata

  • Download URL: algo_secure-0.1.4-py3-none-any.whl
  • Upload date:
  • Size: 103.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.2

File hashes

Hashes for algo_secure-0.1.4-py3-none-any.whl
Algorithm Hash digest
SHA256 b8c9e034bd3ef3be737b7b71514fd7efab45484b305006e93450ae50c8be7391
MD5 82c547571d3432656b4b7b825874c994
BLAKE2b-256 ae2a94bbb1e5701f9f11149ac71cc3a23b837b61c2895f669eeeefc6fed0dc2f

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page