claude-api-ops

Files (20)

• SKILL.md
• .gitkeep · assets/
• agentic-loop.py · assets/
• cached-agent-loop.py · assets/
• output-schema.json · assets/
• recall-probe.py · assets/
• agent-sdk.md · references/
• caching-and-cost.md · references/
• compaction.md · references/
• context-engineering-quickref.md · references/
• context-engineering.md · references/
• messages-api.md · references/
• resources.md · references/
• sdk-examples.md · references/
• structured-outputs.md · references/
• tool-use.md · references/
• .gitkeep · scripts/
• check-model-table.py · scripts/
• context-budget.py · scripts/
• run.sh · tests/

SKILL.md (first lines)

# Claude API Operations

Building applications and agents on Anthropic's API: the Messages API, tool use,
prompt caching, structured outputs, batches, thinking/effort, and the Claude
Agent SDK. For developers writing apps *against* the API — not for using Claude
Code itself.

**API surfaces move fast.** Model IDs, parameters, and betas in this skill were
verified against platform.claude.com (2026-08). When in doubt — especially for
"latest model" or pricing questions — verify with WebFetch against
`https://platform.claude.com/docs/en/about-claude/models/overview.md` or query
the Models API (`client.models.list()`).

## Current Models (verified 2026-08)

| Model | ID (exact, no date suffix) | Context | Max Output | Input $/MTok | Output $/MTok |
|---|---|---|---|---|---|
| Claude Fable 5 | `claude-fable-5` | 1M | 128K | $10.00 | $50.00 |
| Claude Opus 5 | `claude-opus-5` | 1M | 128K | $5.00 | $25.00 |
| Claude Sonnet 5 | `claude-sonnet-5` | 1M | 128K | $2.00 | $10.00 |
| Claude Haiku 4.5 | `claude-haiku-4-5` | 200K | 64K | $1.00 | $5.00 |

Use these alias IDs verbatim. **Never append date suffixes** (`claude-sonnet-5-20260630`
is wrong → 404). Haiku 4.5 is the one current model with a *dated* snapshot id
(`claude-haiku-4-5-20251001`) behind its alias; from the 4.6 generation on, the dateless
id **is** the pinned snapshot.

**Legacy (still available, no longer current):** `claude-opus-4-8`, `claude-opus-4-7`,
`claude-opus-4-6`, `claude-opus-4-5`, `claude-sonnet-4-6`, `claude-sonnet-4-5`. Migrating
off one: `https://platform.claude.com/docs/en/models/opus-5/migration-guide.md` (or run
`/claude-api migrate` in Claude Code). Live capability lookup:
`client.models.retrieve("claude-opus-5")` → `.max_input_tokens`, `.max_tokens`,
`.capabilities` dict.

## Model Selection Decision Tree

```
What is the workload?
│
├─ Hardest problems, long-horizon agents, deep research, ceiling intelligence
│  └─ claude-fable-5 (premium ceiling) or claude-opus-5 (default flagship)
│
├─ Agentic coding, tool-heavy workflows, production assistants
│  └─ claude-opus-5 (quality) or claude-sonnet-5 (speed/cost balance)
│
├─ High-volume production: summarization, RAG answers, extraction
│  └─ claude-sonnet-5
│
├─ Classification, routing, simple Q&A, latency-critical
│  └─ claude-haiku-4-5
│
└─ Subagents inside a larger system
   └─ One tier below the orchestrator (Opus loop → Sonnet/Haiku workers)
```

Tiering rule: route by task difficulty, not by uniform default. An Opus
orchestrator dispatching Haiku classifiers is routinely 5-10x cheaper than
Opus-everywhere with no quality loss on the simple legs.

## Which Surface? (API vs Agent SDK vs Batches)

| Need | Use | Why |
|---|---|---|
| One request → one response (classify, summarize, extract, Q&A) | **Messages API** | Simplest; full control |
| Multi-step pipeline, your code controls the logic | **Messages API + tool use** | You own the loop |
| Custom agent with your own tools, your infra | **Messages API + tool use** (manual loop or SDK tool runner) | Max flexibility |
| Agent that reads/edits files, runs commands, searches — without building tools | **Claude Agent SDK** | Claude Code's tools + agent loop as a library |
| CI/CD automation, coding agents, production agent apps | **Claude Agent SDK** | Built-in tools, hooks, sessions, MCP |
| Large non-urgent workloads (eval runs, backfills, bulk extraction) | **Batches API** | 50% discount, ≤24h turnaround |
| Hosted agent, Anthropic runs loop + sandbox | **Managed Agents** (beta) | No infra; see official docs |

Rule of thumb: start at the simplest tier. Reach for an agent only when the
task is genuinely open-ended (multi-step, hard to fully specify, errors
recoverable, value justifies cost).

## Messages API Quick Start

Everything goes through `POST /v1/messages`. Headers: `x-api-key`,
`anthropic-version: 2023-06-01`, `content-type: application/json`.

```python
# pip install anthropic
import anthropic

client = anthropic.Anthropic()  # reads ANTHROPIC_API_KEY

response = client.messages.create(
    model="claude-opus-5",
    max_tokens=16000,
    system="You are a concise technical assistant.",
    messages=[{"role": "user", "content": "Explain CRDTs in one paragraph."}],
)
for block in response.content:        # content is a list of typed blocks
    if block.type == "text":          # always check .type before .text
        print(block.text)
print(response.stop_reason, response.usage.input_tokens, response.usage.output_tokens)
```

TypeScript and streaming versions: [references/sdk-examples.md](references/sdk-examples.md). Stream long outputs; non-streaming above ~16K `max_tokens` risks SDK HTTP timeouts.

Full params, response shape, stop reasons, errors, retries, rate limits:
[references/messages-api.md](references/messages-api.md)

## Thinking & Effort (quick reference)

- **Adaptive thinking is ON BY DEFAULT on Fable 5 / Opus 5 / Sonnet 5** — send no
  `thinking` field and you still get (and pay for) thinking. On the legacy
  4.6–4.8 models it stays off until you set `thinking: {"type": "adaptive"}`.
- **Manual budgets are gone.** `{"type": "enabled", "budget_tokens": N}` returns a
  **400 on Opus 4.7 and every later model** (Opus 5, Sonnet 5, Fable 5 included);
  deprecated on Opus 4.6 / Sonnet 4.6. Control depth with `effort`, not tokens.
- **Turning thinking off:** Sonnet 5 accepts `{"type": "disabled"}`. Opus 5 accepts
  it only at effort `high` or below — pairing it with `xhigh`/`max` is a **400**.
  Fable 5 **rejects it outright**; thinking there is unconditional, so budget for it.
- **Effort (GA):** `output_config: {"effort": "low" | "medium" | "high" | "xhigh" | "max"}`
  — nested in `output_config`, not top-level. Default `high` (identical to omitting
  it). `xhigh`: Fable 5, Opus 5, Opus 4.8/4.7, **Sonnet 5**. `max`: those plus Opus 4.6
  and Sonnet 4.6. Haiku 4.5 does not support `effort` at all.
- **Sampling params removed on Opus 4.7 and later** (so Opus 5, Sonnet 5, Fable 5):

From 0xDarkMatter/claude-mods, MIT.

More from 0xDarkMatter/claude-mods