CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

jbaruch/speaker-toolkit

Eleven skills for conference speakers covering rhetoric analysis, talk authoring, deck building and export, CFP submissions, publishing, illustrations, shownotes, and screencast production, informed by a 113-entry Presentation Patterns taxonomy (83 observable: 64 patterns + 19 antipatterns; 30 unobservable: 21 patterns + 9 antipatterns).

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

toolkit-root-lookup.mdrules/

alwaysApply:
No
applyTo:
skills/*/SKILL.md, skills/*/references/**/*.md — when writing or reviewing a command that runs a co-shipped toolkit script or reads a co-shipped file
description:
Named lookup rule for the {speaker_toolkit_root} path form toolkit commands use in a clone and an installed mount

Toolkit Root Lookup

Authority Claimed

  • jbaruch/coding-policy: skill-authoring Script References — a resolved plugin root, for skills that run both from a clone and from an installed mount, under this plugin's own named lookup rule.
  • jbaruch/coding-policy: script-delegation Scoped Plugin Contracts — this rule is the named rule its precondition 1 requires.

Lookup

  • Input: the absolute path of the loaded SKILL.md.
  • Output: speaker_toolkit_root, the plugin root two directories above that file's directory, used as an absolute path.
  • The same lookup serves a clone and an installed mount: it reads only the skill's own location.
  • Never derive it from the consumer working directory, PATH, an environment variable, or a git remote.
  • Every shipped SKILL.md carries this lookup.

Authored Scope and Co-Shipped Targets

  • Scope: the skills listed in .tessl-plugin/plugin.json skills, and their references/.
  • Targets: files or directories co-shipped in this plugin under skills/ or rules/, named literally in a command after the resolved root.
  • A directory target is a co-shipped data directory a command passes as an argument (a pattern catalog, a bundled CLI's source tree).
  • Every target, file or directory, resolves inside the plugin root.
  • A target outside this plugin's package source does NOT qualify.
  • Each target script's module docstring (or header comment) is authoritative for its input/output contract.
  • A command's step prose summarizes the inputs, outputs, exit codes and side effects the caller acts on, and leaves internal logic to the script.
  • Consumer-owned paths (talk directories, vault paths, shownotes sites) are never rooted here.

No Authority Granted

  • The lookup selects a path and nothing else.
  • It grants no runtime authority and waives no check after lookup.
  • Each target's own contract, the configured-interpreter rule ({python_path}), and every gate the target names still apply.

Regression Coverage

  • tests/test_installed_plugin_paths.py verifies the lookup and its output:
    • every shipped SKILL.md declares the lookup;
    • the declared lookup resolves the plugin root, from a clone and from an installed mount copy;
    • every rooted command path names a file or directory that exists and resolves inside that root;
    • no operational toolkit path is left unrooted.

README.md

tile.json