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 skills add BLANXLAIT/pantry-agent --skill skillgit clone --depth 1 https://github.com/BLANXLAIT/pantry-agentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/blanxlait/pantry-agent/skill)<a href="https://agentmods.dev/skills/blanxlait/pantry-agent/skill"><img src="https://agentmods.dev/badge/skills/blanxlait/pantry-agent/skill/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/blanxlait/pantry-agent/skill"><img src="https://agentmods.dev/badge/skills/blanxlait/pantry-agent/skill.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00063 | $0.01580 |
| Opus 5.5 | $0.00025 | $0.00632 |
| Sonnet 5.5 | $0.00013 | $0.00316 |
| Haiku 4.5 | $0.00006 | $0.00158 |
Grade A, and why
pantry-agent 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 2mo ago.
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 — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pantry Agent
Grocery shopping assistant powered by the Kroger API. Supports all Kroger-owned banners.
Workflow
- Ask the user for their ZIP code if not already known.
- Find the nearest store using
find_stores. The response includes the chain name (Kroger, Ralphs, etc.) and thelocationId. - Use the
locationIdfrom the store result for all subsequent product searches. - Store, search, and product commands work without authentication.
- Use
preview_cartwith thelocationIdto validate item pricing and availability before adding to cart. No authentication required. - Before adding to cart, call
check_auth_statusto determine if the user is authenticated. - If not authenticated, call
kroger_start_authto get the authorization URL and present it to the user. - After the user completes login, retry the cart or profile operation.
- Always confirm with the user before adding items to their cart.
- Show prices and stock availability when displaying products.
Authentication flow (agent-managed OAuth)
Authentication is decoupled from cart operations — add_to_cart and get_profile return an AUTH_REQUIRED error without opening a browser. Follow these steps when authentication is needed:
- Call
check_auth_statusto see if the user is already authenticated (avoids unnecessary auth prompts). - If not authenticated, call
kroger_start_auth— this starts a local callback server and returns the Kroger authorization URL. - Present the URL to the user and ask them to open it in their browser.
- The user completes the Kroger login in their browser; the callback server captures the token automatically.
- Wait for the user to confirm that login is complete.
- Retry the original cart or profile operation.
Do not attempt to open the URL yourself. Present it to the user for them to open.
You can also use the kroger-oauth prompt to get step-by-step guidance for this flow.
Example agentic flows
Shopping flow (search → preview → add)
1. find_stores(zipCode: "94102") → get locationId
2. search_products(term: "milk", locationId: ...) → pick product UPC
3. preview_cart(items: [{upc, qty}], locationId: ...) → verify price/stock
4. check_auth_status() → verify authentication
5. [if not auth] kroger_start_auth() → user logs in
6. add_to_cart(items: [{upc, qty}]) → confirm to user
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.
- 2mo ago First seen · 141 lines · 63 tokens per session scan A 060bfdd21d58
pantry-agent is a skill published in the GitHub repository BLANXLAIT/pantry-agent (0 stars, last pushed 7mo ago), licensed MIT. It adds 63 tokens to every session and 1,580 once invoked, about $0.0003 per session on Opus 5.5. 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.
pantry-agent compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| pantry-agentBLANXLAIT/pantry-agent | Skill | 0 | 7mo ago | MIT | A | 63 | Claude Code |
| amazon-reviews-api-skillbrowser-act/skills | Skill | 6.3k | 3d ago | MIT | A | 124 | Claude Code, Codex |
| amazon-competitor-analyzerbrowser-act/skills | Skill | 6.3k | 3d ago | MIT | A | 48 | Claude Code, Codex |
| asc-subscription-localizationrorkai/app-store-connect-cli-skills | Skill | 1.1k | today | MIT | A | 60 | Claude Code |
| external-purchasesrshankras/claude-code-apple-skills | Skill | 867 | today | MIT | A | 76 | Claude Code |
| food-orderBitterbot-AI/bitterbot-desktop | Skill | 2.5k | today | MIT | A | 53 | Claude Code, Codex |
| Product Photoshoot Workflowjau123/MeiGen-AI-Design-MCP | Skill | 1.8k | yesterday | MIT | A | 111 | Claude Code |