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/audit-modulegit 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 | $0.00867 |
| Opus 5 | $0.00000 | $0.00434 |
| Sonnet 5 | $0.00000 | $0.00173 |
| Haiku 4.5 | $0.00000 | $0.00087 |
Grade A, and why
audit-module 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.
What it actually says
Audit the rad-modules module: $ARGUMENTS
Resolve $ARGUMENTS to a directory under modules/ (e.g. "Istio_GKE", "Bank_GKE",
"AKS_GKE"). If it doesn't match a directory, report the available modules and stop.
Run every check below and report findings. For each issue, state what is wrong and the
correct fix per .agent/skills/module-conventions/SKILL.md. End with:
"Module $ARGUMENTS: N issue(s) found." or "Module $ARGUMENTS passes all checks."
You may run python3 scripts/check_conventions.py --modules-dir modules/ first for the
mechanical checks, then add the judgement-based ones below.
CHECK 1 — TEN STANDARD VARIABLES
Read the module's variables.tf. Verify all ten standard variables are declared with the correct names: module_description, module_dependency, module_services, credit_cost, require_credit_purchases, enable_purge, public_access, deployment_id, resource_creator_identity, trusted_users. Also verify enable_services is declared.
Note legitimate exceptions: AKS_GKE, EKS_GKE and Migration_Center omit enable_services (they
enable local.default_apis unconditionally); Container_Migration, Migration_Center and
VMware_Engine omit trusted_users. Report
omissions, but classify these as informational for those module types.
CHECK 2 — UIMETA TAGS
Every variable description must contain a {{UIMeta group=N order=M}} tag inside the
description string (not a separate comment). Report any variable missing it. Confirm
enable_services, where present, uses group=0 order=109.
CHECK 3 — SENSITIVE CREDENTIALS
Any variable whose name contains client_secret, aws_secret_key, aws_access_key,
private_key, or service_account_key must set sensitive = true AND must have no
hardcoded default (no default, or default = ""). Report violations.
CHECK 4 — PROVIDER AUTH PATTERN
Determine which pattern the module uses and confirm it is internally consistent:
- Pattern A (direct): a single
provider.tf, no impersonation. Used by AKS_GKE, EKS_GKE. - Pattern B (impersonated):
versions.tf(requirements) +provider-auth.tfdeclaring both thegoogleandgoogle-betaproviders withimpersonate_service_account = length(var.resource_creator_identity) != 0 ? var.resource_creator_identity : null.
Flag: both provider.tf and provider-auth.tf present (should be one or the other); a module
using google-beta under Pattern A (must be Pattern B); a hardcoded access_token.
CHECK 5 — API ENABLEMENT INVARIANT
For every google_project_service resource (usually in main.tf), confirm BOTH:
disable_dependent_services = false
disable_on_destroy = false
and that it does NOT use lifecycle { prevent_destroy = true }. These are critical:
multiple modules share a project, so disabling APIs on destroy breaks the others.
CHECK 6 — LICENSE HEADERS
Every .tf file should begin with the Apache 2.0 block-comment header (Google LLC). Report any .tf file missing it.
CHECK 7 — OUTPUTS
outputs.tf should export at least deployment_id and project_id. For attached-cluster
modules that keep outputs in main.tf, confirm those two outputs exist somewhere. Report if
absent.
CHECK 8 — DOCUMENTATION CONSISTENCY
a) README.md exists and its Inputs table lists every variable in variables.tf with
matching defaults (minus the {{UIMeta}} tag).
b) module_documentation default in variables.tf points to the GitHub URL of
docs/labs/<Module>.md — NOT a LAB_GUIDE.md inside the module directory.
c) The lab guide docs/labs/<Module>.md exists.
d) No LAB_GUIDE.md file exists inside the module directory.
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 · 91 lines · 0 tokens per session scan A d2166e2ae57b
audit-module 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 867 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.