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 agentmods add commands/joncovington/meicagent/paper-reportgit clone --depth 1 https://github.com/joncovington/MEICAgentWhat 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 | $0.00000 | $0.01035 |
| Opus 5 | $0.00000 | $0.00517 |
| Sonnet 5 | $0.00000 | $0.00207 |
| Haiku 4.5 | $0.00000 | $0.00103 |
Grade A, and why
paper-report 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 3d 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 — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Produce the weekly (or custom-range) paper-trading performance report, comparing all four risk profiles side by side. Covers both execution_mode='paper' (forward, live-quote) and execution_mode='replay' (historical SPX) trades in one combined view, since both write to the same data/paper_trades.db schema.
1. Gather the range summary
python src/db.py --db data/paper_trades.db get_range_summary --start <YYYY-MM-DD> --end <YYYY-MM-DD>
Default the range to the last 7 calendar days ending today (ET) unless the user specifies a different window (e.g. "since program start", "last month"). This returns profiles: { "conservative": {...}, "moderate": {...}, "aggressive": {...}, "very-aggressive": {...} }, each with total_trades, win_count/loss_count/win_rate_pct, profit_factor, avg_win/avg_loss, expectancy_per_trade, max_consecutive_losses, max_drawdown, worst_day, net_pnl, and a daily_pnl series (date, net_pnl, cumulative_pnl) per profile.
2. Compute the risk-adjusted suite
get_range_summary gives dollar P&L and drawdown; derive the ratio metrics from each profile's daily_pnl series against the $100,000 virtual bankroll convention:
- Period return for each day =
net_pnl / 100000. - Sharpe = mean(period returns) / stdev(period returns), annualized by
sqrt(252)if daily granularity. - Sortino = same, but the denominator uses only the downside deviation (stdev of negative returns only).
- Calmar = (annualized return) / (max_drawdown / 100000).
- Recovery factor = net_pnl / max_drawdown (skip if max_drawdown is 0).
Flag Sharpe > 3 or profit_factor > 4.0 as a likely overfit/curve-fit warning rather than a stronger pass — note this explicitly in the report rather than presenting it as unqualified good news.
3. Render equity + underwater curves
For each profile, build a markdown sparkline (or a compact ASCII bar sequence) from its daily_pnl.cumulative_pnl series (equity = 100000 + cumulative_pnl) and a parallel underwater curve (running peak − current equity). No plotting dependency needed — this is a text report.
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.
- 3d ago First seen · 51 lines · 0 tokens per session scan A 2e9a7b4e6311
paper-report is a command published in the GitHub repository joncovington/MEICAgent (4 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,035 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 2026-08-31.
Other commands, from other repositories
scan
Universal scan — auto-detects asset class (stock / crypto / index / FX / commodity) and runs the matching protocol. With no args, regenerates scanned/INDEX.md.
scan-flow
Partial scan — refresh ONLY the asset-class flow & positioning section of an existing current.md (Flow & Squeeze for stocks, Crypto/Index/FX/Commodity Flow for non-stocks). Does not re-run charts, 6-pillar, or earnings.
rescan
Refresh existing scan, rotating prior snapshot to archive/. Args: or "macro.
scan-earnings
Earnings setup analysis for ticker. 8-point checklist + pre-print read.
scan-macro
Top-down macro/regime scan. Saves to scanned/MACRO/current.md.
discover
Run a Finviz screener aligned with the framework + current macro regime. Saves a dated candidate list to scanned/SCREENS/.