Borrowing it
Nothing to install: this file belongs to deepgram/cli. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/deepgram/cli/main/.github/prompts/pr-interface-review.mdgit clone --depth 1 https://github.com/deepgram/cliWrote 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/commands/deepgram/cli/pr-interface-review)<a href="https://agentmods.dev/commands/deepgram/cli/pr-interface-review"><img src="https://agentmods.dev/badge/commands/deepgram/cli/pr-interface-review/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/commands/deepgram/cli/pr-interface-review"><img src="https://agentmods.dev/badge/commands/deepgram/cli/pr-interface-review.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 | $0.00000 | $0.02292 |
| Opus 5 | $0.00000 | $0.01146 |
| Sonnet 5 | $0.00000 | $0.00458 |
| Haiku 4.5 | $0.00000 | $0.00229 |
Grade A, and why
pr-interface-review 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 2d 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 — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pull Request Interface & Contract Review
You are reviewing a pull request to assess its impact on the project's public interface and contracts. Your goal is to produce a per-interface review that classifies every signature change by breaking-change risk, identifies what backwards-compatibility measures were taken, and calls out gaps that remain.
This review is strictly about the public surface — types, classes, function signatures, request/response schemas, exported symbols, wire formats, and CLI/config flags. It is not a general code review.
Step 1 — Gather the change set
- Fetch the PR metadata, file list, and diffs. Don't stop at the first page — paginate until you have every changed file.
- Identify the public-surface files. Heuristics:
- Anything re-exported from
__init__.py,index.ts,mod.rs,lib.rs, package entry points, or barrel files. - Files under
types/,models/,schemas/,requests/,responses/,dto/,api/,proto/,openapi/,*.d.ts. - Generated SDK files (Fern, OpenAPI Generator, protoc, etc.) — these are public surface even when auto-generated.
- Any file whose path or name suggests it defines a contract (
*_request.py,*Params,*Response,I*.ts,*.proto).
- Anything re-exported from
- Ignore lockfiles, vendored code, fixtures, internal helpers, and tests — except note when a test was added, deleted, or modified to remove a case (see Step 5).
- If the repo has compat-tracking docs (e.g.
AGENTS.md,CHANGELOG.md,MIGRATION.md,BREAKING.md,.fernignore, codeowner files for "frozen" generated files), read them. They tell you what the maintainers consider stable, what shims already exist, and what's deliberately frozen.
Step 2 — Extract every interface change
For each public-surface file, enumerate every change at the method, class, type-alias, enum, exported-constant, or schema-field level. For each one, record:
- Identifier (fully qualified —
module.Class.method,package::Type, etc.). - Kind: class, function, method, type alias, enum, TypedDict / interface / Pydantic model field, exported symbol, route handler, message schema.
- Before signature/shape (from the
-side of the diff or the base ref). - After signature/shape (from the
+side). - Direction of change: addition, removal, rename, type widening, type narrowing, field add (required vs optional), field remove, default change, reshape (nesting), constructor → alias collapse.
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.
- 2d ago First seen · 121 lines · 0 tokens per session scan A 0c4181f23ff2
pr-interface-review is a command published in the GitHub repository deepgram/cli (10 stars, last pushed 3d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,292 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 2026-09-10.
Other commands, from other repositories
review-branch
Review the current branch's diff against base by dispatching atomic-reviewer. No orchestration loop, no spec required — pre-flight before /commit pr or /commit merge.
done
Finish a task - document, create PR or merge, close.
start
Start session - restore context, show tasks, select work item.
ask
Ask a single expert agent a question - like a Slack ping to a colleague.
note
Add a note to the active task - progress, decisions, context. --session (aliases --snapshot, --pre-compact, --compact) writes a resume-grade session snapshot without ending the session.
statusline
Manage & persist statusline appearance - light/dark, compact, and which rows to show.