Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/techequitycloud/rad-modules/security-scangit clone --depth 1 https://github.com/techequitycloud/rad-modulesWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.01272 |
| Opus 5 | $0.00000 | $0.00636 |
| Sonnet 5 | $0.00000 | $0.00254 |
| Haiku 4.5 | $0.00000 | $0.00127 |
Grade A, and why
security-scan scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Run a security audit of the rad-modules repository: $ARGUMENTS
If $ARGUMENTS names a module (e.g. "Istio_GKE", "EKS_GKE"), scope all checks to that
module directory. Otherwise scan every module under modules/.
This complements the Trivy config scan in CI (which catches generic misconfigurations) by checking the conventions and multi-cloud credential rules specific to rad-modules.
End with the structured summary at the bottom.
CHECK 1 — HARDCODED CLOUD CREDENTIALS
Grep all .tf files in scope for credential fields assigned a literal value: client_secret, aws_secret_key, aws_access_key, aws_secret_access_key, password, api_key, access_token, secret_key, private_key, service_account_key
Flag any assigned a quoted literal (≥8 chars). Safe forms (do NOT flag): = var. = local. = data.<...> = random_password.<...>.result = ""
Also flag any variable "<credential>" block with a non-empty hardcoded default. This is
the cardinal rule: Azure (ARM_*) and AWS (AWS_*) credentials are supplied at runtime via
env vars or tfvars, never baked into the module. Severity: HIGH.
CHECK 2 — SENSITIVE FLAGS
Every credential variable (client_secret, azure_tenant_id, subscription_id, client_id,
aws_secret_key, aws_access_key, aws_secret_access_key, and any *_secret / *_key / password)
must set sensitive = true so the value is redacted in plan output and logs. Report any that
don't. Do NOT flag tenant_id — that is a resource-naming identifier (default "demo"), not a
credential.
Severity: MEDIUM.
CHECK 3 — SERVICE ACCOUNT & IAM LEAST PRIVILEGE
Read every .tf that creates google_project_iam_member, google_project_iam_binding,
google_service_account, or grants roles (gke.tf, iam.tf, hub.tf, etc.).
For each role granted to a service account or to trusted_users, flag anything broader than
needed:
roles/owner — always flag (HIGH)
roles/editor — always flag (HIGH)
roles/iam.securityAdmin, roles/iam.serviceAccountAdmin — flag unless justified (MEDIUM)
roles/container.admin — check whether roles/container.developer suffices (LOW)
roles/storage.admin — check whether objectAdmin/objectViewer suffices (LOW)
Report the SA/binding, the broad role, and a narrower alternative.
CHECK 4 — IMPERSONATION SAFETY
For Pattern B modules (provider-auth.tf present):
a) Both the google and google-beta providers must set
impersonate_service_account = length(var.resource_creator_identity) != 0 ? var.resource_creator_identity : null,
so the module still works under plain ADC when no SA is supplied.
b) No provider or resource may carry a hardcoded access_token.
c) No module currently mints its own token via google_service_account_access_token; if one
is introduced, its lifetime must be bounded (≤ "3600s").
CHECK 5 — API ENABLEMENT INVARIANT
Every google_project_service must set disable_dependent_services = false and
disable_on_destroy = false, and must not use lifecycle { prevent_destroy = true }.
A destroy that disables shared APIs is a denial-of-service against co-tenant modules.
Severity: HIGH (data-integrity / availability).
CHECK 6 — NETWORK EXPOSURE
a) Firewall rules (google_compute_firewall): flag any source_ranges = ["0.0.0.0/0"]
opening management ports (22, 3389, 6443, 10250). Note Istio/LB ingress 80/443 is
usually intentional but call it out.
b) GKE clusters: check whether enable_private_nodes / a private endpoint is configured;
flag clusters with a fully public control plane unless the module's purpose requires it.
c) master_authorized_networks — flag clusters exposing the API server to 0.0.0.0/0.
CHECK 7 — DESTROY-PROVISIONER SAFETY
For null_resource blocks with local-exec provisioners:
a) Destroy provisioners (when = destroy) should use set +e and --ignore-not-found /
|| echo "Warning..." so a partial state never blocks destroy.
b) Triggers must capture every variable the destroy provisioner reads (only
self.triggers.* is available at destroy time) — flag a destroy provisioner that
references var.* directly.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 128 lines · 0 tokens per session scan A 0cc741fd32cf
security-scan is a command published in the GitHub repository techequitycloud/rad-modules (2 stars, last pushed 8d ago), licensed MPL-2.0. It costs nothing until one of its globs matches a file; then it loads 1,272 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.