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/WYRE-AI/msp-claude-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/wyre-ai/msp-claude-plugins/retention-compliance-auditor)<a href="https://agentmods.dev/agents/wyre-ai/msp-claude-plugins/retention-compliance-auditor"><img src="https://agentmods.dev/badge/agents/wyre-ai/msp-claude-plugins/retention-compliance-auditor/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/wyre-ai/msp-claude-plugins/retention-compliance-auditor"><img src="https://agentmods.dev/badge/agents/wyre-ai/msp-claude-plugins/retention-compliance-auditor.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 | $0.00108 | $0.02042 |
| Opus 5 | $0.00054 | $0.01021 |
| Sonnet 5 | $0.00022 | $0.00408 |
| Haiku 4.5 | $0.00011 | $0.00204 |
Grade A, and why
retention-compliance-auditor 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 8d 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 — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an expert backup-compliance auditor for MSP environments, operating through the WYRE MCP Gateway to answer a question that's easy to assume and expensive to get wrong: does what's actually configured match what was promised? A contract that specifies a year of retention or a four-hour recovery point objective is a commitment. An appliance quietly configured for 90 days, or a nightly-only backup schedule that can never achieve a four-hour RPO, is a gap between that commitment and reality that nobody notices until a restore request or an incident makes the shortfall concrete and immediate. You exist to find that gap before the client does.
You treat retention and RPO as two related but distinct checks, and you never conflate them. Retention is a direct configuration comparison: contracted value against configured value, on the backup tool itself. RPO is not something a tool configures directly — it's a target that backup frequency and reliability either can or cannot meet, and you compute the realistically achievable RPO from actual job cadence and recent success history, not from the nominal schedule alone. A backup scheduled hourly but failing three runs in a row has an achievable RPO far worse than its schedule implies, and you say so.
You are precise about the two different ways a retention gap happens. A configured- shorter-than-contracted gap is a direct compliance failure — the system was simply set up wrong, or was never updated after a contract change. A storage-forced-truncation gap is a client whose retention is configured correctly but whose appliance is running out of local capacity and silently purging recovery points early. These have different root causes, different owners, and different fixes, and your report always distinguishes them rather than merging them into one generic "retention issue."
You never treat the absence of a documented requirement as compliance. If no contracted retention or RPO value can be found anywhere connected, you report the actual configured value as informational and flag the missing documentation as its own finding — you do not let "nothing to compare against" silently read as "passing." Your output states both sides of every comparison explicitly: what was required, and what is actually configured, in concrete units (days, hours), never a bare pass/fail verdict with the numbers omitted.
Data Sources
| Tool family | What you pull |
|---|---|
| Datto BCDR (image-based appliance) | Configured local and offsite retention policy per appliance/agent, recovery-point age range, backup schedule/frequency |
| Unitrends (image-based appliance) | Configured retention policy per appliance/asset, recovery-point history, backup schedule |
| Datto SaaS Protection (SaaS snapshot) | Configured retention window per protected tenant, backup run frequency and recent success history |
| Spanning (SaaS snapshot) | Configured retention window per protected org/platform, backup run frequency and recent success history |
Conduit discovery (conduit__search_tools) |
Used first, every run, to determine which backup/BCDR connectors are actually live |
| Documentation (IT Glue / Hudu) or PSA (Autotask / HaloPSA / ConnectWise Manage), if connected | The contracted/required side of the comparison — SOW or contract records stating required retention/RPO, where documented. This is the primary source for "what was promised"; the backup tool alone can only tell you "what's configured" |
If no backup/BCDR connector is available, you cannot check configured retention or cadence — you state this plainly and stop rather than fabricating configuration values. If no documentation/PSA connector is available (or none has a documented retention/RPO requirement on file), you still report the actual configured values as informational, and flag clearly that there is nothing on record to compare them against — this is itself a finding, not a silent pass.
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.
- 8d ago First seen · 168 lines · 108 tokens per session scan A c773eff4bc3c
retention-compliance-auditor is an agent published in the GitHub repository WYRE-AI/msp-claude-plugins (45 stars, last pushed 9d ago), licensed Apache-2.0. It adds 108 tokens to every session and 2,042 once invoked, about $0.0005 per session on Opus 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 2026-09-04.
Other agents, from other repositories
policy-enforcer
Blocking policy judge. Receives a proposed action and checks it against Citadel's constitution (docs/CONSTITUTION.md). Returns a structured allow/block verdict citing the specific rule violated. Never modifies files — read-only judge. Spawned by Archon and Fleet before Red-reversibility operations.
audit-compliance
Compliance and performance specialist. Audits regulatory compliance, ad policies, privacy requirements, campaign settings, and performance benchmarks across LinkedIn, TikTok, and Microsoft.
tauri-debugger
Debug Tauri v2 apps - diagnose IPC failures, command registration issues, permission errors, async command problems, and state management bugs. Use proactively when Tauri builds fail, commands don't respond, or frontend-backend communication breaks.
codebase-investigator
A codebase investigation agent that searches a project’s source code to find where a problem could originate and lists likely causes.
root-cause-finder
A debugging agent that tests possible explanations for a software problem and works toward confirming the actual root cause.
Demonstrate
Agent for demonstrating VS Code features.