verify-before-done

A rule for development work that requires real runtime evidence before calling a fix or feature complete.

In plain words
What is it for?
Use it when changing plugins, authentication, or API integrations to require live startup, OAuth, or endpoint checks before reporting success.
Why use it?
It prevents code review, compilation, or mocked tests from being mistaken for proof that an integration works.

Cursor rule for Cursor

Install

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.

agentmods
npx agentmods add rules/flomenco-inc/flo-plugin/verify-before-done
Clone the repo
git clone --depth 1 https://github.com/Flomenco-Inc/flo-plugin

Made for: Cursor.

Per session 358 This file is loaded in full into every session.
When invoked 358 The same file — it is already loaded in full.
Security scan B 2 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What 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.

ModelPer sessionOnce invoked
Fable 5 $0.00358 $0.00358
Opus 5 $0.00179 $0.00179
Sonnet 5 $0.00072 $0.00072
Haiku 4.5 $0.00036 $0.00036

Measured yesterday against content hash f3926eadb566, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

verify-before-done scanned grade B with 2 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.

Subtle steeringmediumPrompt injection

Instructions that bias recommendations or shape behaviour without the user noticing.

**Never tell the user something is "fixed", "done", or "working" without proof.** The proof must be an actual runtime verification, not a code review of your own changes.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -s -o /dev/null -w "%{http_code}" "https://dev.floapp.co/v1/oauth/google-drive/authorize-url"
.cursor/rules/verify-before-done.mdc · 54 lines

What it actually says

Verify Before Claiming Done

Never tell the user something is "fixed", "done", or "working" without proof. The proof must be an actual runtime verification, not a code review of your own changes.

The rule

Before any of these phrases (or equivalent):

  • "This should now work"
  • "The fix is ready"
  • "Issue is resolved"

You MUST have at least one of the following:

For plugin code changes

Run the plugin locally and confirm it connects:

npm run build && npm start
# Verify: no startup errors, MCP handshake succeeds

For OAuth/auth changes

Test the actual OAuth flow against dev:

# Verify the authorize URL returns a redirect, not a 403/500
curl -s -o /dev/null -w "%{http_code}" "https://dev.floapp.co/v1/oauth/google-drive/authorize-url"

For API integration changes

Call the actual API endpoint the plugin uses:

curl -s -X GET "https://api.dev.floapp.co/v1/<endpoint>" \
  -H "Authorization: Bearer $TOKEN" | jq .

What does NOT count as verification

  • "I've reviewed the code and it looks correct"
  • "The tests pass" (unit tests mock the API)
  • "It compiled successfully"

When the user reports "still broken"

Before making another code change:

  1. Reproduce the exact failure
  2. Follow root-cause-before-fix.mdc before writing code
Changes

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.

  1. yesterday First seen · 54 lines · 358 tokens per session scan B f3926eadb566

Subscribe to this mod's changes

verify-before-done is a cursor rule published in the GitHub repository Flomenco-Inc/flo-plugin (0 stars, last pushed 2d ago), licensed MIT. It adds 358 tokens to every session, about $0.0018 per session on Opus 5. A static security scan graded it B with 2 findings (subtle steering, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.