DeepSeek-native
Built around DeepSeek reasoning, max reasoning effort by default, SSE streaming, tool-use semantics, and prefix-cache behavior. One orca exec hands off the task — no context switch.
GeezerVisonThe World Wide Web, restored to 800 pixels of glory. |
| Home|About|Gaming cabinet|Archives|How it works|Starter sites |
GeezerVison is showing orcaagent.dev — Orca Agent - DeepSeek-native terminal coding agent v0.5.19 · Rust-native Orca Agent — DeepSeek-native terminal coding agentOrca is a local terminal coding agent built around DeepSeek: long-context coding, multi-turn tool use, prefix-cache friendly prompts, resumable history, persistent goals, and approval-aware automation in one Rust binary. 1Mcontext window 99%prefix-cache hit ∞default turn cap 37built-in tools 6platforms orca · TUI — ~/projects/api ›Message Orca… / commands · @ files · $ skills ⇧Tab suggestdeepseek-flash · max · ctx 99% · ? help What you'll notice Built around DeepSeek, not around a generic agent shell.DeepSeek-nativeBuilt around DeepSeek reasoning, max reasoning effort by default, SSE streaming, tool-use semantics, and prefix-cache behavior. One orca exec hands off the task — no context switch. 1M context, self-managedA 1M token window with automatic compaction past the 80% threshold, preserving the system prompt and recent turns. Long tasks keep their context. Persistent goal modeSet a long-running objective with /goal; it auto-continues after each successful turn, survives restarts, and only lets the model complete or block after a goal audit. Approval modesTool specs declare capabilities; reads run directly, while write, shell, network, and agent actions follow your configured approval and sandbox policy. Skills and user inputMarkdown skills can be explicitly injected with $skill ids, and the TUI can answer structured ask_user_question prompts without ending the turn. Rewind and resumeLocal JSONL transcripts support --resume for continuation and --fork for branching; /rewind restores files and conversation to any earlier turn, with model summaries for the dropped side and checkpoints kept in content-addressed storage. Prefix cache Tuned for DeepSeek prefix cache, end-to-end.Nine rounds of real-API tuning. The wire prompt stays append-only at the byte level — system, tools, history, summary baseline — so multi-turn loops, compaction, and resume hold a stable prefix instead of redoing it each turn. 99%post-compaction main-loop hit (real API) 92%non-compacted short-task hit 0duplicate remote summary calls (hashed cache) Quick Start Install once, then just run orca.After installation, start Orca Agent from your terminal. The first run guides you through the DeepSeek API key, then drops you straight into the interactive coding TUI. InstallUse npm for the fastest path, or choose the native curl or PowerShell installer below. npm install -g @blade-ai/orca Run Orca AgentLaunch the TUI. If no API key is configured yet, Orca opens the setup flow and saves it for future sessions. orca Start codingType the task in the interactive terminal. For automated runs later, add verifier commands from the exec workflow. › fix the failing auth test Common dev tasks The loops you already run, with more memory and fewer handoffs.Fix failing testsTrace the failure, edit the right files, run the verifier, and keep the transcript for review. Long refactorsUse persistent goals and automatic continuation when one turn is not enough to land the change. Codebase archaeologySearch, read, and summarize a repo without losing the path from evidence to conclusion. Release checksCombine command execution, JSONL events, and release notes into an auditable final pass. Reusable workflowsRun project or user workflows from .orca/workflows/ when a process deserves a repeatable shape. Approval-aware automationLet reads flow quickly while edits, shell, network, and agent actions respect policy. Control surface Inspect, resume, and gate every turn.From prompt to tool call to result, Orca keeps coding runs readable, verifiable, and resumable instead of hiding them behind a black box. read_fileglobeditgrepbashwrite_filegit_statusweb_searchsubagentWorkflowupdate_planget_goalcreate_goalupdate_goalMCPexternal Sessions & resumeLocal JSONL transcripts under ~/.orca/sessions/, with --resume to continue and --fork to branch a new run. Verifier loopPass --verifier "cargo test" to gate a run on a real command; pass/fail is reported with exit code 2 on failure. Sandbox & hooksPermission profiles can scope filesystem and network access, while lifecycle hooks return structured JSON to deny, modify, or inject context. Structured event stream--output-format jsonl emits readable session, reasoning, approval, tool, workflow, and completion events for automation or audit trails. Why teams pick Orca Agent A local DeepSeek workflow, not another chat window.Regular AI chat
Generic coding agents
Orca Agent
Command surface One set of verbs across the real dev loop.# Hand it a task orca exec "fix this test" # Refactor in full-auto orca exec --approval-mode full-auto "refactor the auth module" # Pick a model + gate on a verifier orca exec --model deepseek-v4-pro --verifier "cargo test" "fix the failing test" ✓ 3 files edited · cargo test passed · exit 0 Technical specs 1M Context window, auto-compacted past the 80% threshold. 6× Native binaries: macOS, Linux, and Windows, arm64 and x64. 16 Built-in, MCP, and external tools share one spec-driven registry. 100% Written in Rust, running as a single local binary. Frequently asked questions Answers before you install.Does Orca Agent only work with DeepSeek?Orca Agent is tuned for DeepSeek V4 and gives the best experience there, but the config can point at an OpenAI-compatible endpoint when your team needs that. Will my session history be lost?No. Orca stores local JSONL transcripts under ~/.orca/sessions/ and supports --resume and --fork. Can I keep control over edits and shell commands?Yes. Built-in, MCP, and external tools declare capabilities, and approval modes plus sandbox profiles decide what can run automatically. What makes persistent goals different from a long prompt?A /goal survives restarts, auto-continues successful turns, and requires an explicit complete or blocked audit before Orca stops. Which platforms does Orca Agent support?Orca Agent ships native binaries for macOS, Linux, and Windows on arm64 and x64, plus an npm package that wraps the binary. How does the sandbox protect my machine?Shell commands run inside a sandbox scoped to the workspace. Linux prefers bubblewrap and falls back to Landlock plus seccomp; macOS uses Seatbelt. Network and writes outside trusted folders need approval. What are approval modes?Approval modes decide how tool calls are gated. Suggest mode asks before writes, auto-edit allows edits but still asks for shell and network, and full-auto runs everything within policy. You can also grant a tool once or for the session. Can Orca Agent use MCP tools?Yes. Orca Agent supports the Model Context Protocol: add stdio or HTTP servers with orca mcp, and their tools appear in the same approval-aware registry as built-in tools. Is Orca Agent the same as other projects named Orca?No. Orca Agent is the DeepSeek-native terminal coding agent in this repository, published at orcaagent.dev. The stablyai/orca desktop app, Plotly's orca package, and the GNOME Orca screen reader are unrelated projects. Install Use npm, or install the native binary directly.npm install -g @blade-ai/orcaSupported platforms: macOS, Linux, and Windows on arm64/x64. Downloads are available on GitHub Releases. |
| You are visitor 01402503 — GeezerVison does not store the pages you look up. Best viewed with Netscape Navigator 4.7 or Internet Explorer 6 at 800×600. No cookies required. |