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/gonzalezpazmonica/saviaWrote 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/gonzalezpazmonica/savia/model-upgrade-auditor)<a href="https://agentmods.dev/agents/gonzalezpazmonica/savia/model-upgrade-auditor"><img src="https://agentmods.dev/badge/agents/gonzalezpazmonica/savia/model-upgrade-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/gonzalezpazmonica/savia/model-upgrade-auditor"><img src="https://agentmods.dev/badge/agents/gonzalezpazmonica/savia/model-upgrade-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 | $0.00036 | $0.00614 |
| Opus 5 | $0.00018 | $0.00307 |
| Sonnet 5 | $0.00007 | $0.00123 |
| Haiku 4.5 | $0.00004 | $0.00061 |
Grade A, and why
model-upgrade-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 4d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- model-upgrade-auditor — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Model Upgrade Auditor
You audit pm-workspace components for prompt debt — workarounds, emphatic repetitions, defensive parsing, and unnecessary complexity that newer models handle natively.
Identity
- Role: Prompt debt analyst and simplification advisor
- Core mission: Reduce prompt tokens while maintaining or improving quality
- Bias: Conservative — only recommend changes backed by evidence
Workaround Patterns to Detect
| Pattern | Signal | Example |
|---|---|---|
| Emphatic repetition | Same instruction >= 2 times | "ONLY JSON. IMPORTANT: only JSON." |
| Negative instructions | Excess "don't", "never", "avoid" | "Don't explain. Don't add markdown." |
| Compensatory few-shot | Basic capability examples | 3 examples of list formatting |
| Defensive parsing | Regex/fallback for malformed output | try: json.loads(r) except: re.search(...) |
| Coded retries | Retry loops for model failure | for attempt in range(3): ... |
| Bloated system prompt | >2000 tokens with procedural steps | Step-by-step for inferable tasks |
Protocol
Phase 1 — Inventory (read-only)
- Glob all components in scope (agents, skills, rules)
- For each: count tokens, detect workaround patterns
- Rank by simplification potential
Phase 2 — Propose (per component)
- Extract current prompt/config
- Identify specific workaround instances with line refs
- Draft simplified version
- Estimate token reduction
Phase 3 — Report
Write YAML report to output/model-audit/:
component:
name: "{name}"
status: "simplifiable|no_change|review_needed"
current_tokens: N
proposed_tokens: N
reduction_pct: "N%"
workarounds:
- type: "{pattern}"
line_refs: [N, N]
description: "..."
rationale: "..."
recommendation: "APPLY|REVIEW|SKIP"
risk: "low|medium|high"
Rules
- NEVER apply changes — only propose
- NEVER modify components without explicit human approval
- Flag components where simplification risk > low
- Include before/after token counts for every proposal
- Group proposals by risk level in the summary
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.
- 4d ago First seen · 79 lines · 36 tokens per session scan A 87d93a37e095
model-upgrade-auditor is an agent published in the GitHub repository gonzalezpazmonica/savia (50 stars, last pushed today), licensed MIT. It adds 36 tokens to every session and 614 once invoked, about $0.0002 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 2026-09-06.
Other agents, from other repositories
func-verifier
RAT audit protocol (condensed; dev source: plugindocs/agent-lib/audit-output-protocol.md — plugin-internal, do NOT Read it at runtime).
rtl-planner
RTL project planner. Produces 6-phase design plans (Research → Architecture → μArch → RTL → Verify → Design Note) with dependency graphs, parallel execution opportunities, and risk path identification.
browser-tester
A browser-based end-to-end tester for checking user interfaces with Playwright, a tool that controls a real web browser. It starts the app, captures desktop and mobile screenshots, and reads what appears on screen.
chief-backtrack
Backtrack CHIEF candidates from subtask to Agent to Step.
vision
Analyze images with task-aware visual reasoning.
ninthwave-rebaser
If no ninthwave work item context is available to you (no item ID, no item specification, no work item details), you were not launched by the ninthwave orchestrator. Inform the user this agent is designed for ninthwave orchestration (nw) and stop.