Muhammad Zakir Ramadhan
9b978fdbe1
Most of this replaces "roughly 550 characters per tool" with the actual per-tool measurements, and fills in the parts a reader hits after the happy path: what a specific error means, what a setting costs, what is not covered. Measured rather than estimated: - Per-tool byte cost, all fourteen, and the per-set totals. 7,673 B for the full set, averaging 548. - Builtin skill bodies at 5,284 B against a 681 B catalogue, which is the argument for loading bodies on demand. - Full system prompt 3,571 chars, core-only 2,045. New sections: - tools: which sets to keep and why, the jail function itself, an output-cap table, and the real error strings for edit_file and multi_edit. - configuration: env var per provider preset, cost-estimate limits, what each --no-* flag isolates, and three settings that do more than they look like. - agents: step caps per variant, which variant to reach for, and the fact that reasoning is charged as output and discarded first by compaction. - headless: exit code 0 means "the turn completed", not "the answer was yes" — with the jq pattern for gating on content. Timeouts, concurrent -c runs fighting over one session, CI recipes for --no-skills. - mcp: parallel connect, startup cost, a debugging ladder, and that toolSets does not gate MCP tools. - registry: publishing, local testing over http://localhost, and a troubleshooting section keyed on the actual validator messages. - memory: what compaction discards in what order, /compact versus automatic pruning, and that -c matches on cwd. - skills: the frontmatter reader's limits, and how to verify a skill loaded. Corrections found while cross-checking against the source: - The guard table was missing --force-with-lease and > /dev/sd… - The done event's token fields are optional, so the jq example filters on one rather than assuming it. Two honest limits now written down: the guard matches command strings, so a base64-decoded or script-wrapped command is not caught; and a registry index is trusted for its contents, not its authorship. Verified: all internal links and heading anchors resolve, every docs/ page is reachable from the README, 538 tests pass, typecheck clean.
Shiro Neko
An agentic coding CLI. It reads your code, edits it, runs your tests, and asks when the request is ambiguous — in a terminal UI, with every mutating action gated behind an approval prompt.
Install
A single prebuilt binary. No runtime, no node_modules.
# macOS, Linux
curl -fsSL https://raw.githubusercontent.com/zakirkun/shiro-neko/main/scripts/install.sh | sh
# Windows
irm https://raw.githubusercontent.com/zakirkun/shiro-neko/main/scripts/install.ps1 | iex
Both verify the download against the release checksums before installing. Builds are
published for linux-x64, linux-arm64, darwin-x64, darwin-arm64, and windows-x64.
Or from source:
git clone https://github.com/zakirkun/shiro-neko
cd shiro-neko
bun install
bun run install:local # builds and puts `shiro` on PATH
First run
shiro
With no API key configured it opens provider setup: pick an endpoint, paste a key, choose
from the models that endpoint actually reports. Settings land in
~/.shiro-neko/config.json. Run /provider any time to change them.
shiro-neko 0.1.0-beta.3 openai/gpt-5 session 0193ab2c
agent: default thinking: medium
cwd: /home/you/project
skills: debug, refactor, review, test
plugins: guard, time
approvals: on for write_file, edit_file, multi_edit, bash, mcp__*
/help for commands
> why does the pagination test fail?
What it does
Answers about your code, grounded in your code. grep goes through ripgrep when it is
installed and honours .gitignore. list_dir gives an ignore-aware tree so it stops globbing
blindly to orient, and read_many_files pulls a batch in one round trip. read_file refuses
binaries rather than filling the context with mojibake.
Edits with your approval. Every write_file, edit_file, multi_edit, and bash call
stops for a y/a/n decision, with a coloured diff for edits. multi_edit is atomic, so
a failing match leaves the file untouched rather than half-changed. The guard plugin
refuses irreversible commands outright — rm -rf, git reset --hard, force pushes,
DROP TABLE — and --yolo cannot bypass it.
Shows its work. Reasoning streams to a collapsed panel you can expand with ctrl-r, the
tool in flight is named as it runs, and bash output streams live instead of arriving all at
once when the command exits. ctrl-c kills a runaway command without ending the turn.
Takes prompts while it works. Type during a turn and it queues; the queue drains in order
when the turn ends. esc interrupts and clears it. @ completes workspace paths.
Reads git without touching it. git_status, git_diff, git_log, git_show, and
git_blame are approval-free, because they spawn git with a fixed argument list and cannot
mutate anything.
Asks instead of guessing. When a request has two readings that lead to different work, the agent puts a question on screen with options.
Delegates searches. task spawns a read-only subagent whose findings come back as one
message, so a search across forty files does not fill the main context. Its progress
streams to a panel.
Extensible from the prompt. /registry browses external skills and plugins and installs
them with one confirmation. A skill is shown in full before its text joins your system prompt;
a plugin is a manifest of refusal rules, never code.
Remembers between sessions. Decisions, working commands, and traps go into per-project memory that is injected at the start of every future session.
Survives long tasks. The task list and project memory live outside the message array,
so they survive both automatic pruning and /compact.
Runs headless. shiro -p "review this diff" --json for scripts and CI.
Keeps the tool list affordable. Fourteen built-in tools, grouped into sets. Each costs
about 550 characters of schema on every request, so { "toolSets": [] } trims back to the six
core ones and a disabled set reaches neither the wire nor the prompt.
Documentation
Start with whichever question you have. Each guide says what it decided and why, not just what the flags are.
| Guide | Contents |
|---|---|
| Configuration | config file, provider presets, environment, every flag |
| Tools | every tool, tool sets and what they cost, the approval model |
| Agents and thinking | variants, thinking levels, step caps, which to reach for |
| Skills | the bundled skills, writing your own, why the catalogue is split |
| Plugins | the interface, the guard and its limits, builtin versus installed |
| Registry | installing external skills and plugins, publishing your own |
| Memory and state | memory, task lists, sessions, compaction and its repair |
| MCP | connecting servers, namespacing, cost, debugging one |
| Headless mode | -p, JSON events, exit codes, CI recipes |
| Architecture | how the loop works and why it is built this way |
| Development | building, testing, adding a tool, releasing |
| Roadmap | what is next and what has been declined |
| TODO | the current work list, with known rough edges |
Commands
Type / and a menu appears, narrowing as you type.
/help /agent [name] /think [level] /provider /models /model <id>
/skills /plugins /registry [search|add|remove] /init /context
/todos /notes /memory /tools /compact /cost
/sessions /resume <id> /save /clear /exit
esc dismisses a panel, interrupts a running turn, and clears the queue. ctrl-c kills the
running command but keeps the turn. ctrl-r expands the reasoning panel. @ completes a
workspace path. Up and down recall earlier prompts.
Status
Working: the agent loop, tool approvals, subagents, skills, plugins, per-project memory,
session persistence, MCP, markdown rendering, headless mode, five-platform builds, streaming
reasoning display, the mid-turn prompt queue, gateable tool sets, read-only git tools, batch
reads, @file completion, interruptible commands, and the external registry.
Next up is in TODO.md; the longer view and what has been declined are in
ROADMAP.md. The short version of what is missing: a summary of what compaction
discarded, web_fetch, a spend ceiling, and a cheaper model for subagent searches.
License
MIT. See LICENSE.