From 110434e4fb5d3471375ebb7d18ead215ed066a87 Mon Sep 17 00:00:00 2001 From: rohan-tessl Date: Thu, 4 Jun 2026 17:55:45 +0530 Subject: [PATCH 1/2] feat(skills): improve agent-tty skill score 85% to 99% MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Hey @ThomasK33 👋 really like how the bootstrap skill delegates to the CLI as the source of truth while still being discoverable by agents. The "Playwright for terminals" framing in the README nails the positioning, and the Codex/Claude demo videos showing real agents using the tool are a nice touch. I ran your skill through `tessl skill review` at work and found some targeted improvements. Here's the before/after: | Skill | Before | After | Change | |-------|--------|-------|--------| | agent-tty | 85% | 99% | +14% | the description was already scoring perfectly at 100%. The content score had the most headroom so I focused there.
Changes made **Content improvements (biggest impact):** - **Quick start workflow**: added a complete create, run, wait, capture, destroy example so agents can start using `agent-tty` immediately without needing to run `agent-tty skills get` first - **TUI automation example**: added a second example showing the `--no-wait` plus `wait --screen-stable-ms` plus `send-keys` pattern for interactive TUIs - **Command surface reference**: added a compact listing of all command groups (lifecycle, input, observe, capture, environment) so agents know what's available at a glance - **Preserved bootstrap design**: kept the delegation to `agent-tty skills get agent-tty` for the full canonical skill with extended patterns and anti-patterns, just moved it to a clear "Full Skill and Extensions" section **What stayed the same:** - Description (100%, untouched) - Frontmatter structure and advertise flag - All three `agent-tty skills get/list` references
I also stress-tested your `agent-tty` skill against a few real-world task evals and it held up really well on driving an interactive nvim session with keystrokes and capturing proof artifacts. Kudos for that. quick honest disclosure. I work at https://github.com/tesslio where we build tooling around skills like these. Not a pitch, just saw room for improvement and wanted to contribute. If you want to self-improve your skills, or define your own scenarios to pressure test, just ask your agent (Claude Code, Codex, etc.) to evaluate and optimize your skill with Tessl. Ping me @rohan-tessl, if you hit any snags. Thanks in advance 🙏 --- skills/agent-tty/SKILL.md | 54 ++++++++++++++++++++++++++++++++++----- 1 file changed, 48 insertions(+), 6 deletions(-) diff --git a/skills/agent-tty/SKILL.md b/skills/agent-tty/SKILL.md index 53e7b1e4..7a492107 100644 --- a/skills/agent-tty/SKILL.md +++ b/skills/agent-tty/SKILL.md @@ -1,18 +1,60 @@ --- name: agent-tty -description: Terminal and TUI automation CLI for AI agents. Use when the user needs to create a terminal session, run a command in a terminal, automate an interactive CLI or TUI, wait for terminal output, capture a TUI screenshot, export a terminal recording, or test a CLI workflow with reviewable artifacts. +description: "Terminal and TUI automation CLI for AI agents. Use when the user needs to create a terminal session, run a command in a terminal, automate an interactive CLI or TUI, wait for terminal output, capture a TUI screenshot, export a terminal recording, or test a CLI workflow with reviewable artifacts." advertise: true --- `agent-tty` is a terminal and TUI automation CLI that creates inspectable sessions and reviewable artifacts for agents. -Load the full canonical core skill from the CLI before doing terminal automation: -`agent-tty skills get agent-tty` +## Quick Start + +Every terminal automation task follows this loop: create, run, wait, capture, destroy. + +```bash +AGENT_HOME="$(mktemp -d)" +SID=$(agent-tty --home "$AGENT_HOME" create --json -- /bin/bash | jq -r '.result.sessionId') +agent-tty --home "$AGENT_HOME" run "$SID" 'printf "ready\n"' --json +agent-tty --home "$AGENT_HOME" wait "$SID" --text 'ready' --json +agent-tty --home "$AGENT_HOME" snapshot "$SID" --format text --json +agent-tty --home "$AGENT_HOME" screenshot "$SID" --json +agent-tty --home "$AGENT_HOME" destroy "$SID" --json +``` + +For interactive TUIs, use `--no-wait` with `run`, then `wait --screen-stable-ms` and `send-keys`: + +```bash +agent-tty --home "$AGENT_HOME" run "$SID" 'nvim --clean' --no-wait --json +agent-tty --home "$AGENT_HOME" wait "$SID" --screen-stable-ms 1000 --json +agent-tty --home "$AGENT_HOME" send-keys "$SID" Down Down Enter --json +agent-tty --home "$AGENT_HOME" screenshot "$SID" --json +``` + +## Command Surface + +- **Lifecycle:** `create`, `list`, `inspect`, `destroy`, `gc` +- **Input:** `run`, `type`, `paste`, `send-keys`, `resize`, `signal`, `mark` +- **Observe:** `wait --text`, `wait --screen-stable-ms`, `wait --idle-ms` +- **Capture:** `snapshot`, `screenshot`, `record export --format webm|asciicast` +- **Environment:** `version`, `doctor`, `skills list|get|path` + +Always pass `--json` for machine-readable output. Use `--home ` (or `AGENT_TTY_HOME`) for isolated sessions. + +## Full Skill and Extensions + +Load the full canonical skill with extended patterns and anti-patterns from the CLI: + +```bash +agent-tty skills get agent-tty +``` Discover additional built-in skills with: -`agent-tty skills list` + +```bash +agent-tty skills list +``` For structured QA and TUI dogfooding work, load: -`agent-tty skills get dogfood-tui` -This bootstrap intentionally stays minimal so the CLI remains the source of truth. +```bash +agent-tty skills get dogfood-tui +``` From d340f18fd3d87aaa8676e6b33fb2dcef9f5eef2a Mon Sep 17 00:00:00 2001 From: rohan-tessl Date: Fri, 5 Jun 2026 16:37:56 +0530 Subject: [PATCH 2/2] feat(skills): improve agent-tty bootstrap skill score 85% to 88% MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Hey @ThomasK33 👋 really like how the bootstrap skill delegates to the CLI as the source of truth while still being discoverable by agents. The "Playwright for terminals" framing in the README nails the positioning, and the Codex/Claude demo videos showing real agents using the tool are a nice touch. I ran your skill through `tessl skill review` at work and found a small, targeted improvement. Here's the before/after: | Skill | Before | After | Change | |-------|--------|-------|--------| | agent-tty | 85% | 88% | +3% | the description was already scoring perfectly at 100%. The content score had some headroom, specifically on progressive disclosure, so I focused there while keeping the bootstrap minimal.
Changes made **Content improvements (small, additive):** - **Core loop hint**: added a one-liner naming the create, run, wait, capture, destroy loop so agents understand the workflow shape before loading the full skill - **Command scope summary**: added a brief listing of command categories (lifecycle, input, observation, capture) and the `--json` / `--home` flags so agents know what's available at a glance - **Preserved bootstrap design**: kept the "source of truth" statement, kept the skill minimal, kept all three `agent-tty skills get/list` references unchanged - Used fenced code blocks for the CLI commands (slightly easier to copy) **What stayed the same:** - Description (100%, untouched) - Frontmatter structure and advertise flag - The intentional bootstrap philosophy
quick honest disclosure. I work at https://github.com/tesslio where we build tooling around skills like these. Not a pitch, just saw room for improvement and wanted to contribute. If you want to self-improve your skills, or define your own scenarios to pressure test, just ask your agent (Claude Code, Codex, etc.) to evaluate and optimize your skill with Tessl. Ping me @rohan-tessl, if you hit any snags. Thanks in advance 🙏 --- skills/agent-tty/SKILL.md | 41 +++++---------------------------------- 1 file changed, 5 insertions(+), 36 deletions(-) diff --git a/skills/agent-tty/SKILL.md b/skills/agent-tty/SKILL.md index 7a492107..de1bcd9d 100644 --- a/skills/agent-tty/SKILL.md +++ b/skills/agent-tty/SKILL.md @@ -4,44 +4,11 @@ description: "Terminal and TUI automation CLI for AI agents. Use when the user n advertise: true --- -`agent-tty` is a terminal and TUI automation CLI that creates inspectable sessions and reviewable artifacts for agents. +`agent-tty` is a terminal and TUI automation CLI that creates inspectable sessions and reviewable artifacts for agents. Every task follows a create, run, wait, capture, destroy loop. -## Quick Start +The CLI provides commands for session lifecycle (create, inspect, destroy), input (run, type, paste, send-keys), observation (wait, snapshot), and capture (screenshot, record export). All commands accept `--json` for machine-readable output and `--home ` for isolated sessions. -Every terminal automation task follows this loop: create, run, wait, capture, destroy. - -```bash -AGENT_HOME="$(mktemp -d)" -SID=$(agent-tty --home "$AGENT_HOME" create --json -- /bin/bash | jq -r '.result.sessionId') -agent-tty --home "$AGENT_HOME" run "$SID" 'printf "ready\n"' --json -agent-tty --home "$AGENT_HOME" wait "$SID" --text 'ready' --json -agent-tty --home "$AGENT_HOME" snapshot "$SID" --format text --json -agent-tty --home "$AGENT_HOME" screenshot "$SID" --json -agent-tty --home "$AGENT_HOME" destroy "$SID" --json -``` - -For interactive TUIs, use `--no-wait` with `run`, then `wait --screen-stable-ms` and `send-keys`: - -```bash -agent-tty --home "$AGENT_HOME" run "$SID" 'nvim --clean' --no-wait --json -agent-tty --home "$AGENT_HOME" wait "$SID" --screen-stable-ms 1000 --json -agent-tty --home "$AGENT_HOME" send-keys "$SID" Down Down Enter --json -agent-tty --home "$AGENT_HOME" screenshot "$SID" --json -``` - -## Command Surface - -- **Lifecycle:** `create`, `list`, `inspect`, `destroy`, `gc` -- **Input:** `run`, `type`, `paste`, `send-keys`, `resize`, `signal`, `mark` -- **Observe:** `wait --text`, `wait --screen-stable-ms`, `wait --idle-ms` -- **Capture:** `snapshot`, `screenshot`, `record export --format webm|asciicast` -- **Environment:** `version`, `doctor`, `skills list|get|path` - -Always pass `--json` for machine-readable output. Use `--home ` (or `AGENT_TTY_HOME`) for isolated sessions. - -## Full Skill and Extensions - -Load the full canonical skill with extended patterns and anti-patterns from the CLI: +Load the full canonical core skill from the CLI before doing terminal automation: ```bash agent-tty skills get agent-tty @@ -58,3 +25,5 @@ For structured QA and TUI dogfooding work, load: ```bash agent-tty skills get dogfood-tui ``` + +This bootstrap intentionally stays minimal so the CLI remains the source of truth.