Claude Skills by thgMatajs

The base rules must apply to every response, and skills are model-invoked — Claude decides per turn whether a skill is relevant, so no `description` wording makes one fire reliably every time. Intermittent formatting is the worst possible outcome here: a user who cannot trust the shape of the output has to read defensively, which is the cost the plugin exists to remove. Output styles modify the system prompt directly and apply to every turn, and `force-for-plugin: true` applies ours automatic...

Search the user's cross-project hoard for what they already recorded about a subject: past corrections, decisions, bugs and their fixes. Only for an explicit /squirrel:dig invocation.

Restructure a rambling ticket, email, pasted note, file, or Jira issue - prose the user received - into the fixed digest brief (TL;DR, Next action, Breakdown, Priority). Trigger on an explicit /squirrel:digest invocation, or an explicit request to digest or restructure a named piece of prose into that brief. An ordinary-language question like 'what should I do with this?' triggers this only when what was pasted alongside it is itself recognisably a ticket, an email, or a written note - a trac...

Run the squirrel-mode calibration interview: seven one-at-a-time multiple-choice questions that build the user's personal response-formatting profile. Only for an explicit /squirrel:init invocation.

Turn squirrel-mode's formatting off for the rest of the current session only. Only for an explicit /squirrel:off invocation.

Turn squirrel-mode's formatting back on for the current session after /squirrel:off. Only for an explicit /squirrel:on invocation.

Resume work on this project after an interruption or at the start of a new session: show recent wins, what was being done, the next action, and open decisions, in that order, then stop. Trigger only on an explicit /squirrel:pickup invocation, or an explicit request to resume or pick up this project's past work at the start of a new session or after a stated interruption. Never trigger on a request to continue the live, current conversation (e.g. 'let's continue' or 'where were we' about what ...

Converge a messy, explicitly given idea dump - unstructured, in need of scoping - into a scoped action plan with one first action and a parking lot for tangents. Trigger only on an explicit /squirrel:plan invocation, or an explicit request to turn a rambling, not-yet-scoped idea dump into that plan. Never trigger on an already-scoped request (e.g. planning the rollout of a defined feature across environments), and never on ordinary architecture questions or design discussion.

Manually load the squirrel-mode base rules (only needed if the forced output style has been disabled).

Record one durable memory in the user's cross-project hoard: a correction, a decision with its reasoning, a bug and its fix, or a fact worth keeping. Only for an explicit /squirrel:stash invocation.

Change one field of an existing squirrel-mode profile without repeating the calibration interview. Only for an explicit /squirrel:tune invocation.

Restructure a rambling ticket, email, pasted note, file, or Jira issue - prose the user received - into the fixed digest brief (TL;DR, Next action, Breakdown, Priority). Trigger on or an explicit request to digest or restructure a named piece of prose into that brief. An ordinary-language question like 'what should I do with this?' triggers this only when what was pasted alongside it is itself recognisably a ticket, an email, or a written note - a tracker key or URL, mail headers, or a messag...

Run the squirrel-mode calibration interview: seven one-at-a-time multiple-choice questions that build the user's personal response-formatting profile. Only run this when the user explicitly asks to run squirrel-mode calibration - for example, asked to run squirrel init or calibrate squirrel-mode. Never start this interview unprompted, and never run it merely because no profile exists yet.

Converge a messy, explicitly given idea dump - unstructured, in need of scoping - into a scoped action plan with one first action and a parking lot for tangents. Trigger only on an explicit request to turn a rambling, not-yet-scoped idea dump into that plan. Never trigger on an already-scoped request (e.g. planning the rollout of a defined feature across environments), and never on ordinary architecture questions or design discussion.

Change one field of an existing squirrel-mode profile without repeating the calibration interview. Only run this when the user explicitly asks to change or tune a squirrel-mode profile field. Never change the profile unprompted, and never run it merely because the user mentioned a preference in passing without asking for the change.

Restructure a rambling ticket, email, pasted note, file, or Jira issue - prose the user received - into the fixed digest brief (TL;DR, Next action, Breakdown, Priority). Trigger on or an explicit request to digest or restructure a named piece of prose into that brief. An ordinary-language question like 'what should I do with this?' triggers this only when what was pasted alongside it is itself recognisably a ticket, an email, or a written note - a tracker key or URL, mail headers, or a messag...

Converge a messy, explicitly given idea dump - unstructured, in need of scoping - into a scoped action plan with one first action and a parking lot for tangents. Trigger only on an explicit request to turn a rambling, not-yet-scoped idea dump into that plan. Never trigger on an already-scoped request (e.g. planning the rollout of a defined feature across environments), and never on ordinary architecture questions or design discussion.