budget

Included with Lifetime

$97 forever

Use to set + track token / cost budgets, get warnings on overrun, and optimise efficiency. Backed by `Budget` class in src/services/tokens.ts. Pairs with /siftcoder:budget command.

General

What this skill does


# budget

Token + cost awareness. Set a limit; track usage; surface warnings; optimise.

## Sub-modes

- **set <limit>** — set the budget limit (tokens or USD)
- **status** — show current consumption
- **report** — historical usage by skill / agent / session
- **optimise** — recommend savings (model downsizing, prompt caching, compression)

## Method (set)

1. Limit can be **tokens** (int) or **USD** (float w/ pricing assumption).
2. Scope: **session** (default) | **daily** | **monthly** | **per-task**.
3. Persist to `~/.siftcoder/budget.json`.
4. Warn at 80%, hard-stop at 100% (configurable).

## Method (status)

1. Read consumption ledger
2. Show: limit / used / remaining / pct / projected (linear extrapolation if mid-window)

## Method (report)

1. Aggregate from memory: tokens-in / tokens-out per summary.
2. Group by skill / agent / model.
3. Top consumers ranked.

## Method (optimise)

1. Find high-consumers (top 5 by cost).
2. For each:
   - Could it use a smaller model? (Haiku vs Sonnet vs Opus)
   - Could prompt caching reduce repeat cost?
   - Could output compression reduce tokens-out?
   - Is there an Ollama-eligible substitute?
3. Estimate savings.

## Output shape (status)

```
Budget:           1,000,000 tokens (session)
Used:             423,450 (42%)
Remaining:        576,550
Projected:        780,000 by session end (78%)
Warning at:       80%
Hard-stop at:     100%
```

## Output shape (optimise)

```
Top consumers (this session):
  /siftcoder:reverse-prompt   142,000 tokens   $0.42
  /siftcoder:archaeologist     98,000 tokens   $0.29
  /siftcoder:codemap           54,000 tokens   $0.16

Recommendations:
  1. reverse-prompt → use cache (8 of 12 invocations had identical fingerprint)
     Estimated savings: 110k tokens / $0.32

  2. archaeologist → switch to Haiku for first pass; only escalate to Sonnet if confidence < 0.7
     Estimated savings: 45k tokens / $0.13

  3. codemap → enable Ollama for embedding-only ops
     Estimated savings: 30k tokens / $0.09

Total projected savings: 185k tokens / $0.54 (~44% of session)
```

## Rules

- **Hard-stop is honoured.** No exceptions without explicit user override.
- **Warnings non-blocking** — status update only.
- **Use real measurement.** `gpt-tokenizer` for token counts; current Anthropic pricing for USD.
- **Per-skill granularity** for report — don't lump everything together.
- **Memory captures budget events** so trends emerge over weeks.

## Anti-patterns

- Setting a budget then ignoring the warning
- USD-based limits without naming the model (different price points)
- Optimising before measuring (premature optimisation)
- Treating soft warnings as hard-stops (alert fatigue)

## When NOT to use

- Single-shot scripted task — overhead > value
- Local-only Ollama work — token cost is electricity, not API
- Pre-launch crunch — wrong time to start counting

## Subagent dispatch

- None — local computation
- Memory MCP for the consumption ledger
- Use `src/services/tokens.ts` `Budget` class directly

## Value over native CC

CC has no native budget tracking. The Budget primitive + skill provide the discipline. Useful for long autonomous runs (`/autonomous`) where token spend can grow unboundedly.

Files: 1

Size: 3.4 KB

Complexity: 7/100

Category: General

Source: https://github.com/ialameh/sift-coder/tree/main/skills/meta/budget

Related in General

modeling-omnistudio-epc-catalog

Included

Salesforce Industries CME EPC product-modeling skill for Product2-based catalog creation. Use when creating EPC products, configuring product attributes, building offer bundles with Product Child Items, or reviewing EPC DataPack JSON metadata for product catalog changes. TRIGGER when: user creates or updates Product2 EPC records, AttributeAssignment payloads, AttributeMetadata/AttributeDefaultValues, Offer bundles, or ProductChildItem relationships. DO NOT TRIGGER when: designing OmniScripts/FlexCards/Integration Procedures (use building-omnistudio-omniscript, building-omnistudio-flexcard, or building-omnistudio-integration-procedure), implementing Apex business logic (use generating-apex), or troubleshooting deployment pipelines (use deploying-metadata).

Generalscripts

relationship-science-coach

Included

Use this skill for direct, practical adult relationship coaching: couples conflict, repair, trust, marriage, dating, flirting, attachment patterns, emotional connection, sex, desire differences, eroticism, kink negotiation, affection, love languages, breakups, and long-term passion. Draw on Gottman, EFT and Hold Me Tight, attachment science, modern sex research, Perel, Nagoski, Kerner, Schnarch, Love and Stosny, and flexible love-language tools. Be concrete and low-hedge. Redirect only for imminent danger, abuse, coercive control, minors, non-consent, self-harm, stalking, or medical/legal/psychiatric decisions.

Generalscripts

building-sf-integrations

Included

Salesforce integration architecture and runtime plumbing with 120-point scoring. Use this skill to set up Named Credentials, External Credentials, External Services, REST/SOAP callout patterns, Platform Events, and Change Data Capture. TRIGGER when: user sets up Named Credentials, External Services, REST/SOAP callouts, Platform Events, CDC, or touches .namedCredential-meta.xml files. DO NOT TRIGGER when: Connected App/OAuth config (use configuring-connected-apps), Apex-only logic (use generating-apex), or data import/export (use handling-sf-data).

Generalscripts

venue-templates

Included

Access comprehensive LaTeX templates, formatting requirements, and submission guidelines for major scientific publication venues (Nature, Science, PLOS, IEEE, ACM), academic conferences (NeurIPS, ICML, CVPR, CHI), research posters, and grant proposals (NSF, NIH, DOE, DARPA). This skill should be used when preparing manuscripts for journal submission, conference papers, research posters, or grant proposals and need venue-specific formatting requirements and templates.

Generalscripts

let-fate-decide

Included

Draws the 12 Houses of the Zodiac Tarot spread to inject entropy into planning when prompts are vague, ambiguous, or casually delegated. Interprets the spread to guide next steps. Use when the user says 'let fate decide', 'YOLO', 'whatever', 'idk', or other nonchalant phrases, makes Yu-Gi-Oh references, or when you are about to arbitrarily pick between multiple reasonable approaches. Prefer over ask-questions-if-underspecified when the user's tone is casual or playful rather than precision-seeking.

Generalscripts

net-ops

Included

Cross-platform network troubleshooting (Windows, macOS, Linux) via local or remote shell. Use for: DNS broken, can't resolve hostnames, nslookup/dig works but apps fail, NRPT, WFP, scutil, /etc/resolver, systemd-resolved, /etc/resolv.conf, NetworkManager, VPN DNS leak residue (ProtonVPN/Mullvad/WireGuard/AnyConnect), AV/firewall blocking DNS or DoH, Tailscale DNS interaction, intermittent connectivity, remote diagnostics over SSH.

Generalscripts