Ship-Safe Bundle: 20 Offline Checks for Dev & Data Teams
The problem
Every one of these failures is cheap to catch before release and expensive afterwards: a key in git history, a migration that locks a table, an API change your consumers find first, a license you can't ship, an import that half-loads, a ledger that no longer balances. Each skill below is a script your agent runs and reads back to you.
See what you're buying
Every skill's own page shows a real command, the real input and the real output, copied from a run. Open any of them from the bundle page and check that it does what it says. Want to try one first? Secrets Leak Scanner is free on its own, and it's included here as a bonus.
What's included (list price each)
Security and supply chain
- Secrets Leak Scanner: Find leaked keys before you push (free, included as a bonus)
- Prompt Injection Scanner: Vet text before agents read it ($12)
- GitHub Actions Security Audit: Find risky workflows ($12)
- Dockerfile Linter: Security and image-size fixes ($9)
- Dependency License Audit: Flag AGPL and unknown licenses ($12)
- File Manifest Checker: Prove a folder came back unchanged ($12)
Backend and data safety
- SQL Migration Safety Check: Catch locking DDL pre-deploy ($15)
- OpenAPI Breaking Changes: Diff two specs before release ($15)
- CSV Import Validator: Exact line and column of bad rows ($12)
- RelayBox Webhook Debugger: Why the HMAC check failed ($12)
- Schema Change Review: Approve or block with lineage risk ($12)
- Ledger Integrity Check: Find drift before you migrate ($12)
LLM evals and evidence
- LLM Eval Regression Check: Compare saved runs offline ($15)
- Search Ranking Eval: nDCG, MRR, recall regression gate ($12)
- Citation Checker: Find unsupported claims in AI answers ($9)
- A/B Test Stats: Sample size, p-value and SRM check ($12)
Business metrics and finance
- SaaS MRR Audit: Reconcile invoices before approving ($12)
- Cap Table Dilution Model: Ownership after each round ($12)
- Release Notes from Commits: Draft and flag bad messages ($9)
- WCAG Contrast Audit: Find failing color pairs and fixes ($12)
Bought separately: $228 at list price. Bundle price: $79 (save 65%). That is less than any 8 of the paid skills bought one at a time at list price (the 8 cheapest come to $87).
Honest limits
These are checks, not guarantees. Each page lists what its script doesn't catch and how it was tested. Most evidence is public data or synthetic test data, and each page says which.
How it works
You own every skill in the bundle individually. Your agent loads each one like any other skill you own. Bundles are sold on the Agensi website; agents can't buy them.
Price and guarantee
$79 one time. Covered by Agensi's money-back guarantee if the skills don't do what their pages promise.
Works with any agent that signs in to Agensi, such as Claude, ChatGPT, Claude Code, Cursor and Codex CLI.
One payment, lifetime access. 20 skills unlock instantly in your library.
What's included
20 skillsWhat will you own after the seed round, with the option-pool top-up and SAFEs and notes converting? Model each round and get price per share, new shares and ownership for every holder, with the math checks shown.
Your AI answer cites a source, but does the source say it? Checks every sentence against its cited passages and flags missing citations, weak support and numbers that aren't in the source, so you review the risky lines first.
Your muted caption text looks fine to you and fails WCAG for everyone else. Check every foreground/background pair in your design tokens against AA or AAA and get the nearest color that passes.
The import fails halfway, or worse, succeeds with bad data. Check a CSV against your column rules and get the exact row, column and code for every bad value, duplicate key and short row.
Find the AGPL package before your lawyer does. Classify every dependency license in a package-lock as allow, review, deny or unknown against your own policy, with an exit code for CI.
Find the Dockerfile lines that slow your builds or widen your attack surface: unpinned bases, apt caches, root user, a COPY that busts the cache. Each finding comes with a line number and a fix.
Did the folder come back from the build, the zip or the transfer exactly as it left? Create a SHA-256 manifest and verify any copy: it lists modified, missing and extra files.
One pull_request_target with write-all permissions can hand over your repo. Audit GitHub Actions workflows for unpinned actions, excessive permissions, script injection and secrets echoed to logs, with the line number for each.
Before you migrate a ledger, prove the books still balance. Finds unbalanced transactions, bad reversals and stored balances that drifted from the entries, with the transaction for each.
A column rename looks harmless until the downstream tables and dashboards break. Turn your lineage facts into a risk score and an APPROVE / CONDITIONAL / BLOCK decision, with the conditions to meet before merge.
Don't learn about a breaking change from your API consumers. Diff two OpenAPI 3 JSON specs and list every break: removed endpoints, new required fields, narrowed enums and changed responses.
Is that conversion lift real, or noise? Plan sample size before the test, then get the p-value, confidence interval and a sample-ratio-mismatch warning from your counts. Plain Python, no stats library.
Your agent reads web pages, READMEs and tool output, and any of them can carry hidden instructions. Scan text first for instruction overrides, spoofed roles, secret-access and exfiltration cues, with the evidence masked.
For RelayBox webhooks: your receiver returns 401 and you can't see why. Recompute the HMAC from the raw body, check the timestamp window and see which step failed. Also finds duplicate deliveries in a log. Not for Stripe, GitHub or Slack, whose own libraries already verify signatures.
Turn a commit log into a changelog draft and see which messages broke the format first. Groups conventional commits, calls out breaking changes and lists non-conventional ones for you to fix.
Your MRR number and your invoices disagree and you don't know why. Rebuild MRR from subscription events, flag invoice price mismatches and see the month-by-month bridge before a board or finance review.
Did your ranking change make search better or worse? Compare a candidate run against relevance labels and a baseline: nDCG, MRR, precision and recall per query, the worst queries listed, and a pass/fail gate for CI.
Find the API key before it's in git history. Scans a folder for leaked OpenAI, Anthropic, AWS, GitHub, Stripe and other keys, redacts the values in the report, and separates real-looking secrets from placeholders so you rotate the right ones first.
A one-line migration can lock a production table. Scan Postgres and MySQL migration files for blocking DDL, risky NOT NULL adds and non-concurrent indexes, with a safer alternative for each.
Your prompt tweak fixed one case and quietly broke three. Compare a saved eval run against a baseline, offline, and see exactly which cases regressed, with pass rate and cost. Exact, contains and JSON-subset checks.
See it in action
Cap Table Dilution Model: Ownership after each round
Example session with this skill installed
Model the illustrative seed round with a cap-priced SAFE, a discounted interest-bearing note and a 15% post-round reserve. Show Markdown and fail if aggregate founders fall below 70%.
- Read your context and instructions
- Compiled the cap table dilution
Cap table dilution model
Illustrative only; not legal/financial advice.
Gate: FAIL
Convention: pre-money cap; conversion shares excluded from pricing capitalization.
example-seed
Pre-money: 4000000.00; post-money: 5000000.00
Round price/share: 3.51944139; effective paid/share: 3.5194413942619027
Pool top-up: 136544 shares; pool after: 15.000057%
Total shares: 1576954; founder ownership: 57.072052%
Unallocated cash from whole-share rounding: 0.00000000
| Holder | Kind | Shares | Before % | After % | Dilution pp | Relative dilution % |
|---|---|---|---|---|---|---|
| example-founder | founder | 900000 | 90.000000 | 57.072052 | 32.927948 | 36.586609 |
| example-pool | pool | 236544 | 10.000000 | 15.000057 | -5.000057 | -50.000571 |
| example-safe | other | 113654 | 0.000000 | 7.207185 | -7.207185 | n/a |
| example-note | other | 42620 | 0.000000 | 2.702679 | -2.702679 | n/a |
| example-investor | other | 284136 | 0.000000 | 18.018027 | -18.018027 | n/a |
Conversion example-safe: cap, price 1.75972070, shares 113654, effective price 1.75972689, unallocated 0.70388828
Conversion example-note: discount, price 2.81555312, shares 42620, effective price 2.81557954, unallocated 1.12622125
Checks: ownership sum 100%; integer shares PASS; gate FAIL
The command exits 1 because the 70% founder policy fails; ownership and integer-share invariants pass. The founder holds 900,000 shares throughout. Confirm the conversion convention before comparing alternative raises or reserve targets.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads these skills and uses them on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every skill in this bundle goes through our trust checks.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.