Build cost tracking and budget enforcement #97

New Issue

Timmy · 2026-03-30T15:40:27Z

Timmy commented

2026-03-30 15:40:27 +00:00

Objective

Track every cloud API call's cost and enforce daily/monthly budgets. Timmy should know exactly how much cognition he's renting and stay within Alexander's budget.

Parent Epic

#94 — Grand Timmy: The Uniwizard

Implementation

Cost Tracking

Every cloud call logs:

{
    "timestamp": "2026-04-01T10:30:00Z",
    "backend": "claude",
    "model": "claude-sonnet-4-20250514",
    "prompt_tokens": 1500,
    "completion_tokens": 800,
    "cost_usd": 0.0069,
    "task_id": "issue-94",
    "task_type": "reasoning",
    "duration_ms": 3200
}

Budget Enforcement

budgets:
  daily_limit_usd: 5.00
  monthly_limit_usd: 50.00
  per_backend:
    claude: 20.00/month
    kimi: 15.00/month
    gpt: 10.00/month
    gemini: 5.00/month
  warning_threshold: 0.8  # warn at 80% spend
  hard_stop: true  # refuse cloud calls at limit

When Budget Exceeded

Timmy falls back to local for everything
Posts a notice: "Cloud budget exhausted, operating local-only until {reset_date}"
Never silently fails

In Evennia

costs command shows today's spend, month's spend, per-backend
budget command shows limits and remaining
Observatory room displays cost dashboard
Alert Script warns when approaching limits

Deliverables

agent/cost_tracker.py — per-call cost logging
agent/budget.py — enforcement logic
configs/budgets.yaml — budget configuration
commands/costs.py — Evennia commands
SQLite table for cost history

Acceptance Criteria

Every cloud call logged with cost
Daily and monthly totals accurate
Hard stop works (blocks cloud calls at limit)
Timmy degrades to local-only gracefully
Cost history queryable (show me last week's spend)

## Objective Track every cloud API call's cost and enforce daily/monthly budgets. Timmy should know exactly how much cognition he's renting and stay within Alexander's budget. ## Parent Epic #94 — Grand Timmy: The Uniwizard ## Implementation ### Cost Tracking Every cloud call logs: ```json { "timestamp": "2026-04-01T10:30:00Z", "backend": "claude", "model": "claude-sonnet-4-20250514", "prompt_tokens": 1500, "completion_tokens": 800, "cost_usd": 0.0069, "task_id": "issue-94", "task_type": "reasoning", "duration_ms": 3200 } ``` ### Budget Enforcement ```yaml budgets: daily_limit_usd: 5.00 monthly_limit_usd: 50.00 per_backend: claude: 20.00/month kimi: 15.00/month gpt: 10.00/month gemini: 5.00/month warning_threshold: 0.8 # warn at 80% spend hard_stop: true # refuse cloud calls at limit ``` ### When Budget Exceeded - Timmy falls back to local for everything - Posts a notice: "Cloud budget exhausted, operating local-only until {reset_date}" - Never silently fails ### In Evennia - `costs` command shows today's spend, month's spend, per-backend - `budget` command shows limits and remaining - Observatory room displays cost dashboard - Alert Script warns when approaching limits ## Deliverables - `agent/cost_tracker.py` — per-call cost logging - `agent/budget.py` — enforcement logic - `configs/budgets.yaml` — budget configuration - `commands/costs.py` — Evennia commands - SQLite table for cost history ## Acceptance Criteria - [ ] Every cloud call logged with cost - [ ] Daily and monthly totals accurate - [ ] Hard stop works (blocks cloud calls at limit) - [ ] Timmy degrades to local-only gracefully - [ ] Cost history queryable (show me last week's spend)

ezra was assigned by Timmy

2026-03-30 15:40:27 +00:00

Timmy referenced this issue

2026-03-30 15:41:44 +00:00

[POLICY] Google AI Ultra leverage boundaries — media/research sidecar only #28

Rockachopa referenced this issue

2026-03-30 15:48:45 +00:00

[RESEARCH] Multi-Backend LLM Routing Landscape — Uniwizard Design Input #101

Rockachopa referenced this issue

2026-03-30 15:50:21 +00:00

[EPIC] Grand Timmy — The Uniwizard #94

Rockachopa commented

2026-03-30 15:50:32 +00:00

Reframe: Budget Enforcement → Quota Utilization Tracking

Per Alexander's directive: the money is already spent. ~$500/month across backends. Unused quota = wasted money.

This issue should flip from "don't overspend" to "are we USING what we're paying for?"

Alerts should fire on:

Backend at <50% utilization mid-cycle → route more work there
Backend approaching 0% remaining → gracefully shift load, don't panic
Total utilization across all backends <70% → Timmy is being too conservative

The ceiling matters. The floor matters MORE.

## Reframe: Budget Enforcement → Quota Utilization Tracking Per Alexander's directive: the money is already spent. ~$500/month across backends. Unused quota = wasted money. This issue should flip from "don't overspend" to "are we USING what we're paying for?" Alerts should fire on: - Backend at <50% utilization mid-cycle → route more work there - Backend approaching 0% remaining → gracefully shift load, don't panic - Total utilization across all backends <70% → Timmy is being too conservative The ceiling matters. The floor matters MORE.

Timmy referenced this issue

2026-03-30 15:58:50 +00:00

[EPIC] Grand Timmy — The Uniwizard #94

Timmy commented

2026-03-30 16:03:23 +00:00

Role Transition

Timmy now owns execution — building, coding, implementing.
Ezra moves to persistent online ops — monitoring, triage, review, cron, 24/7 watchkeeping.

Timmy: this is yours. Read the ticket, build it, PR it. Ezra reviews.

Timmy — build cost tracking. Log every cloud API call with token counts and cost. Enforce daily/monthly budgets. Fall back to local when budget exhausted.

## Role Transition **Timmy** now owns execution — building, coding, implementing. **Ezra** moves to persistent online ops — monitoring, triage, review, cron, 24/7 watchkeeping. Timmy: this is yours. Read the ticket, build it, PR it. Ezra reviews. Timmy — build cost tracking. Log every cloud API call with token counts and cost. Enforce daily/monthly budgets. Fall back to local when budget exhausted.

ezra was unassigned by Timmy

2026-03-30 16:03:23 +00:00

Timmy self-assigned this 2026-03-30 16:03:23 +00:00

ezra referenced this issue

2026-03-31 16:30:05 +00:00

[EPIC] Claude Code Source Study — Reference Architecture for Grand Timmy #154

ezra referenced this issue

2026-03-31 16:30:10 +00:00

[STUDY] Cost tracking — learn from Claude Code's usage pricing system #158

Sign in to join this conversation.