cost-projection — Claude Code skill for spend forecasts
The cost-projection Claude Code skill projects future spend from recent session data. It estimates daily costs and shows when configured budget thresholds may be reached.
MIT 249 installs ★ 82,234 updated 2026-10-25
npx skills add ruvnet/ruflo --skill cost-projection -a claude-codeWhat cost-projection does
The skill reads session records from the cost-tracking namespace, measures spend over a chosen window, then calculates a daily USD rate. It extends that rate across selected horizons, with 7d, 30d, 90d and 365d as the defaults. Results come as a table or JSON.
When a budget has been set with cost-budget set, the output estimates time to the 75%, 90% and 100% thresholds and marks thresholds already passed. The skill uses session-* records and the Bash tool. Its linear forecast assumes the current spend rate continues, so a workload shift calls for a fresh run.
What it helps with
- project spend across a chosen time horizon
- estimate when spending will reach a budget threshold
- send a spend forecast to a budget dashboard as JSON
- check how soon the budget may be fully consumed
Example prompts for the agent
- Project spend for the last 7d across a 30d horizon and return JSON.
- Estimate the days remaining until the 75%, 90% and 100% budget thresholds.
- Show projected spend for 7d, 30d and 90d in a table.
How to install cost-projection
1Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill cost-projection -a claude-code2Manually: put the files in place
~/.claude/skills/cost-projection/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/cost-projection/
1Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill cost-projection -a codex2Manually: put the files in place
.agents/skills/cost-projection/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill cost-projection -a cursor2Manually: put the files in place
.cursor/skills/cost-projection/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install ruvnet/ruflo plugins/ruflo-cost-tracker/skills/cost-projection2Manually: put the files in place
.github/skills/cost-projection/SKILL.md1Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill cost-projection -a gemini-cli2Platform CLI
gemini skills install https://github.com/ruvnet/ruflo --path plugins/ruflo-cost-tracker/skills/cost-projectionManually: put the files in place
.devin/skills/cost-projection/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill cost-projection -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about cost-projection
How do I install cost-projection in Claude Code?
Run npx skills add ruvnet/ruflo --skill cost-projection -a claude-code, then open ~/.claude/skills/cost-projection/SKILL.md.
What does cost-projection do?
It derives a daily spend rate from recent session records and extends that rate across selected horizons. With a configured budget, it also estimates when spending will reach set thresholds.
How does cost-projection handle changes in spend rate?
It uses a linear projection based on the current rate. Run it again after a workload shift to recalculate.
Source text
SKILL.md · MIT
--- name: cost-projection description: Forward-looking spend extrapolation. Computes a USD-per-day rate from the recent measurement window, projects to 7d/30d/90d/365d horizons, and surfaces "days until budget exhausted" when a budget is configured. Predictive counterpart to `cost-budget-check` (reactive). argument-hint: "[--window 7d] [--horizons 7d,30d,90d,365d] [--format table|json]" allowed-tools: Bash disable-model-invocation: true --- Forward-looking cost projection that pairs with `cost-budget-check`: - **`cost-budget-check`** answers "have we crossed the line?" — reactive - **`cost-projection`** answers "when will we cross the line?" — predictive ## Algorithm 1. Read all `session-*` records from the `cost-tracking` namespace (same source `cost-budget-check` uses). 2. Filter to a measurement window (default last 7 days on `capturedAt`/`startedAt`). 3. Compute the per-day burn rate: `windowSpend / windowDays`. 4. Linear-extrapolate to each requested horizon (default 7d/30d/90d/365d). 5. If a budget is configured (`budget-config` key set by `cost-budget set`): - Compute "days until 75% / 90% / 100% consumed" at the current rate. - Surface as a separate table block with `ALREADY REACHED` markers for thresholds already exceeded. ## Flags --window <duration> Measurement window. `Nh|Nd|Nw|Nm`. Default `7d`. --horizons <csv> Projection horizons. Default `7d,30d,90d,365d`. --format table|json Default `table` (markdown). Env: `PROJECTION_NAMESPACE` override (default `cost-tracking`), `PROJECTION_QUIET=1` (alias for `--format json`). ## Smoke transcript (3 sessions × $1 over 7d, $20 budget) ``` | Sessions in window | 3 | | Window spend | $3.000000 | | **USD per day** | **$0.428571** | | All-time spend | $3.000000 across 3 sessions | ## Projected spend (linear extrapolation) | Horizon | Days | Projected spend | |---|---:|---:| | 7d | 7 | $3.0000 | | 30d | 30 | $12.8571 | ## Budget exhaustion ($20.00 configured) | Threshold | Target | Remaining | Time at current rate | |---|---:|---:|---| | 75% | $15.00 | $12.00 | 28.0 days | | 90% | $18.00 | $15.00 | 35.0 days | | 100% | $20.00 | $17.00 | 39.7 days | ``` ## When to use - **Finance / SRE planning**: hand the JSON output to a budget dashboard for "are we on track for the quarter?". - **CI gates**: `cost-projection --format json | jq '.budget.exhaustion[2].daysUntilReached < 7'` → fail builds when 100% exhaustion is < 1 week away. - **Post-workload-shift sanity check**: after a big feature lands, re-run to verify the rate hasn't accelerated past expectations. ## Stationarity assumption Linear extrapolation assumes the current rate holds. The footer reminds operators to re-run after workload shifts. For drift detection over MULTIPLE windows, pair with `cost-trend` (which already covers benchmark-drift).
© ruvnet · MIT · ruvnet/ruflo