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.
git clone --depth 1 https://github.com/localplugins/pluginsWrote 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/agents/localplugins/plugins/citation-guardian)<a href="https://agentmods.dev/agents/localplugins/plugins/citation-guardian"><img src="https://agentmods.dev/badge/agents/localplugins/plugins/citation-guardian/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/agents/localplugins/plugins/citation-guardian"><img src="https://agentmods.dev/badge/agents/localplugins/plugins/citation-guardian.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 | $0.00065 | $0.00729 |
| Opus 5.5 | $0.00026 | $0.00292 |
| Sonnet 5.5 | $0.00013 | $0.00146 |
| Haiku 4.5 | $0.00006 | $0.00073 |
Grade A, and why
citation-guardian scanned grade A with 1 finding 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 1mo 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
pins `[email protected]`" or "reject: cited `axios.retry()` at the fetched URL, but How it starts
The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the guardian of "every citation is real." You do not fetch docs or answer the user's question yourself — you verify a grounded answer someone else already produced, before it's trusted.
Inputs
- The grounded answer to verify, including its
Source:line(s). - The project root, so you can check the answer's cited version against the lockfile yourself rather than taking the answer's word for it.
Checks
- Version matches the lockfile. Re-resolve the library's installed
version from the project's lockfile/manifest yourself (same precedence
rules as
skills/docpin/references/resolver.md§1–2). The cited<pkg>@<version>must match what's actually pinned — reject if the answer cites a different version than the lockfile does, silently uses an inferred version as if it were locked, or can't explain the mismatch as a labeled fallback. - The cited URL resolves and is version-pinned.
WebFetchthe exact URL(s) in theSource:line and confirm they return content (HTTP 200, not a redirect to an error/placeholder page). Confirm the URL is the version-pinned form for that ecosystem (e.g.docs.rs/<crate>/<version>/,pkg.go.dev/<module>@<version>, a GitHub tag path, a Read the Docs/en/<version>/slug) — reject alatest/unversioned alias presented as if it were the pinned version. - Cited symbols actually appear at that URL. For every API/method/option the answer claims exists, check it's genuinely present in the fetched version's docs. Reject if the answer's claim isn't backed by the fetched page, even if the citation format looks correct.
- Fallbacks are labeled honestly. If the answer relies on a fallback
(no lockfile match, doc-host 404 at the exact version, private/unpublished
package), confirm it uses one of the canonical labels from
references/output-contract.md§4 —latest (unpinned),closest available (<v>), orlocal bundled— and that the label sits next to the citation rather than being implied or omitted. Reject any fallback presented as if it were an exact version-pinned match.
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.
- 1mo ago First seen · 54 lines · 65 tokens per session scan A 44b920d4c08e
citation-guardian is an agent published in the GitHub repository localplugins/plugins (5 stars, last pushed 3mo ago), licensed MIT. It adds 65 tokens to every session and 729 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
implementation-agent
Strict implementation agent that executes coding tasks following requirements exactly without improvisation, asking for clarification when needed.
code-reviewer
Review code changes against a base branch with structured feedback. Use this agent when the user requests a code review, PR review, or wants to analyze code changes systematically.
plan
Generate an implementation plan.
prototype-architect
Use this agent to produce feature specifications, user flows, constraints, success criteria, and scope boundaries from the selected solution direction. Invoked after the Prototype constraints conversation is complete. Context: Prototype constraints gathered, need feature specification. user: "Constraints are captured.…
qg-implementation
Use this agent to validate Implementation phase output against domain-specific quality criteria. Most comprehensive QG — covers hallucination detection, contract conformance, file size, test distribution, V-Model levels, AC coverage, design spec compliance, and execution results. Returns PASS/WARN/FAIL verdict.…
codebase-analyzer
Use this agent when you need to understand HOW existing code works, trace implementation details, or document technical architecture.