New here? Try asking your AI to "help me plan" or "review my code" -- IJFW handles the rest. Memory and costs appear here as you work.
Today's Cost
$981
5 sessions, 5 projects
Claude, measured from API tokens
Active Session
ijfw
claude-opus-4-7 -- current
Cache Efficiency
97.0%
2.71B tokens served from cache (30d)
Claude Code caches automatically. IJFW measures your efficiency.
30-Day Actual
$6,546
Claude measured
Today by Project
| Project | Platform | Spend |
|---|---|---|
| ijfw | Claude | $392 |
| wayland | Claude | $241 |
| aion | Claude | $223 |
| socialmedia | Claude | $124 |
| others | $1 |
Live Feed Real data from ~/.ijfw/observations.jsonl (50 entries)
15:37recalldecisions -- session_start (mcp)
15:19recalldecisions -- session_start (mcp)
15:16recallsession_start -- decisions (mcp)
15:15recallsession_start -- decisions (mcp)
15:05recalldecisions -- session_start (mcp)
14:59recallsession_start -- decisions (mcp)
14:58recalldecisions -- session_start (mcp)
13:48recalldecisions -- session_start (mcp)
13:47recalldecisions -- session_start (mcp)
13:46recalldecisions -- session_start (mcp)
13:33recalldecisions -- session_start (mcp)
13:29recalldecisions -- session_start (mcp)
Showing memory-recall events. Tool call and file edit events populate as sessions run.
Usage Limit
Account labels are optional -- add one to track multiple subscriptions.
Usage window shown as an estimate based on today's token spend vs. your daily average.
Spend by Project (30 days, Claude measured)
| Project | Last Active | 30d Cost | % |
|---|---|---|---|
| ijfw | 2026-04-16 | $2,736 | 42% |
| wayland | 2026-04-16 | $2,126 | 32% |
| Flow | 2026-03-27 | $662 | 10% |
| socialmedia | 2026-04-14 | $605 | 9% |
| Crucible | 2026-03-25 | $110 | 2% |
| hearth | 2026-03-30 | $89 | 1% |
| PiPLLM | 2026-03-18 | $45 | <1% |
| openclaw | 2026-04-14 | $45 | <1% |
| mindgraph | 2026-04-08 | $40 | <1% |
| aion | 2026-04-16 | $23 | <1% |
| ContentDev | 2026-04-07 | $19 | <1% |
| dev | 2026-03-31 | $13 | <1% |
| pip | 2026-04-11 | $10 | <1% |
| twitterclean | 2026-04-05 | $8 | <1% |
| deepcode | 2026-03-26 | $7 | <1% |
| TestBEd | 2026-04-13 | $5 | <1% |
| tradecanyon-sentinal | 2026-03-24 | $4 | <1% |
Claude cost is measured precisely from API tokens. Codex and Gemini costs are flat-rate subscriptions -- shown separately for clarity.
By Model (Claude measured, 30d)
| Model | 30d Cost | Cache hit | Sessions |
|---|---|---|---|
| Opus 4.6 | $6,544 | 97.0% | -- |
| Sonnet 4.6 | $2 | -- | -- |
| Codex and Gemini are flat-rate subscriptions -- shown separately from per-token Claude costs. | |||
Opus 4.6 is your primary model -- 99.9% of Claude cost. Cache hit rate 97.0% is measured from 2.71B cache_read tokens over 30 days.
Activity Types
Activity tracking activates as you work. Types and one-shot rates appear here.
Activity classification populates automatically from observed tool usage.
30-Day Cost Trend (Claude measured)
Mar 17Total: $6,546 Claude (30d measured)Apr 16
Claude measured cost only. Codex/Gemini subscriptions ($270/mo combined) are flat-rate and excluded from this trend.
Memory Search -- all 35 files
/
Project Memory (35)
Donahoe Loop pattern170 recalls
Positive Framing is Sacred feedback12 recalls
Memory Tiers project8 recalls
Brainstorm is a Conversation feedback6 recalls
Cross-AI Combo Policy project5 recalls
Workflow Owns Execution feedback4 recalls
Cross-Audit UX project4 recalls
Ownership Discipline feedback3 recalls
Always Recommend and Show Progress feedback3 recalls
Dual-Mode Workflow project2 recalls
Workflow Clarity and Navigation feedback2 recalls
No Proxy Rule project2 recalls
Hook Event Semantics project2 recalls
Cross-Audit Must Auto-Fire feedback1 recall
Claude-Side Specialist Swarm feedback1 recall
Phase 7 Scope project1 recall
IJFW Forge Portability project1 recall
Installer Must Merge project1 recall
Plugin Auto-Register MCP project1 recall
Prelude Tool project1 recall
FTS5 Deferred project1 recall
Live Todo Visibility feedback1 recall
Don't Make Me Think Cross-Audit feedback1 recall
Book-Tool Recursion project1 recall
Session State 2026-04-14 project1 recall
Migration Philosophy feedback1 recall
Mockups Must Be Real HTML feedback1 recall
Auto-Invoke Design Skills feedback1 recall
Design Requires Approval Before Build feedback1 recall
Prompt Improver project1 recall
Cross-Project Search project1 recall
Never Publish Without Live E2E feedback1 recall
No Hanging Items feedback1 recall
Reports Must Be Self-Contained feedback1 recall
IJFW Design Dispatcher project1 recall
Donahoe Loop
patterncoreIJFW
Donahoe Loop -- canonical IJFW loop: audit gates between every stage, Multi-AI Trident handoff at each gate.
Plan -- Audit -- Execute -- Verify -- Ship, with cross-AI review at each gate.
Positive Framing is Sacred
feedbackUX
Positive framing is sacred -- no negatives ever in user-facing surfaces. Sutherland reframe rule: reposition every message as opportunity.
--
Open in editor to view full file content.
Sessions (48 in 30 days)
| Date | Project | Model | Trident |
|---|---|---|---|
| 2026-04-16 | ijfw | opus-4-6 | active |
| 2026-04-16 | wayland | opus-4-6 | -- |
| 2026-04-16 | aion | opus-4-6 | -- |
| 2026-04-16 | socialmedia | opus-4-6 | -- |
| 2026-04-15 | ijfw | opus-4-6 | -- |
| 48 total sessions over 30 days. | |||
Handoff Documents
No handoffs yet. Handoffs are created automatically at session end -- start a coding session to see them here.
Handoff documents appear here as you create them with /handoff.
All Projects (17 with spend, Claude measured 30d)
| Project | Last Active | 30d Cost | Model |
|---|---|---|---|
| ijfw current | 2026-04-16 | $2,736 | Opus 4.6 |
| wayland | 2026-04-16 | $2,126 | Opus 4.6 |
| Flow | 2026-03-27 | $662 | Opus 4.6 |
| socialmedia | 2026-04-14 | $605 | Opus 4.6 |
| Crucible | 2026-03-25 | $110 | Opus 4.6 |
| hearth | 2026-03-30 | $89 | Opus 4.6 |
| PiPLLM | 2026-03-18 | $45 | Opus 4.6 |
| openclaw | 2026-04-14 | $45 | Opus 4.6 |
| mindgraph | 2026-04-08 | $40 | Opus 4.6 |
| aion | 2026-04-16 | $23 | Opus 4.6 |
| ContentDev | 2026-04-07 | $19 | Opus 4.6 |
| dev | 2026-03-31 | $13 | Opus 4.6 |
| pip | 2026-04-11 | $10 | Opus 4.6 |
| twitterclean | 2026-04-05 | $8 | Opus 4.6 |
| deepcode | 2026-03-26 | $7 | Opus 4.6 |
| TestBEd | 2026-04-13 | $5 | Opus 4.6 |
| tradecanyon-sentinal | 2026-03-24 | $4 | Opus 4.6 |
Cross-AI Review History
Clean slate this week. Run /cross-audit to generate your first cross-AI review.
Cross-AI reviews get a second opinion from other AI engines -- findings appear here.
Extensions
Loading extensions...
Active Extension
Loading...
Audit Charts
Events / hour (24h)
Deny rate by extension
Top denied tools
Quota usage
Permission Events
No permission events yet.
Platform Subscriptions
Claude Max 20x$200/mo
Codex Pro$20/mo
Gemini AI Ultra$249.99/mo
Combined flat-rate$469.99/mo
Billing day1st of month
Prices pinned2026-04-16
IJFW uses your subscription tier to calculate accurate cost and savings projections. No tier configured? Defaults work fine.
Account Labels
Labels let you track which subscription generated which sessions.
Account labels are optional. Most users never need them. Add one if you rotate between multiple Max accounts.
Display
Light mode
CurrencyUSD
Platform filterAll platforms
DensityComfortable
Cost methodologyView methodology
Data Confidence
Claude costMeasured
Cache hit rateMeasured
Codex costFlat-rate
Gemini costFlat-rate