terminal-feature-work

name	terminal-feature-work
description	Terminal emulator feature implementation workflow for this Kotlin repository. Use when adding or changing terminal protocols, parser/core/host behavior, modes, SGR, OSC, CSI, DCS, Unicode, input, rendering contracts, or feature-gap documentation.

Terminal Feature Work

Use this skill for any new terminal protocol or behavior in this repository.

Workflow

1. Identify the owning layer:
   • ketraterm-parser recognizes byte/protocol syntax and emits semantic commands.
   • ketraterm-core owns grid physics, modes, width, attributes, storage, and state.
   • ketraterm-host maps existing parser semantics to existing core APIs.
2. Read the relevant module AGENTS.md.
3. Add tests that describe real terminal behavior before or alongside code.
4. Implement the smallest layer-correct behavior.
5. Add full-path tests when bytes should flow through the parser.
6. Update docs/terminal-feature-map.md and docs/terminal-feature-gap-map.md when support or scope changes.

Rules

• Do not fake unsupported behavior.
• Do not move width, cursor bounds, grid mutation, or mode persistence into the parser.
• Do not parse ANSI/OSC/SGR in host.
• Keep hot paths primitive, table-shaped, and allocation-conscious.
• Prefer existing module helpers and fixture style.
• When implementing or extending query/response features (such as DECRQSS or XTGETTCAP), or when creating new terminal features that can be queried, always update the security allowlist of queried settings or capabilities in the response channel (TerminalResponseChannel), and reject unauthorized or unsupported queries with standard protocol-defined failure responses.

Done

The work is done only when normal, omitted/default, malformed, hostile, and boundary cases are tested; unsupported parts are explicit TODOs; and the parser/core/host boundary remains clean.