Clarify p.glob and addon composition guidance (#248)

* Initial plan

* Clarify prompt glob and addon guidance

Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>

Author: Copilot

skills/rig/SKILL.md

@@ -47,7 +47,7 @@ Defaults: `name: "agent"`, `model: "small"`, `maxTurns: 4`, string input/output,
 | Several known files | `p.readAll(paths)` |
 | Static shell command | `p.bash(command)`; use ``p.bashRaw`...` `` for literal backslashes |
 | Caller-supplied path(s) | `p.readInput(field)` / `p.readAllInput(field)` with `s.path` schemas |
-| Discover workspace paths | `p.glob(pattern)` (prefer over `p.bash("find ...")` for static discovery) |
+| Discover workspace paths | `p.glob(pattern)` returns paths only; follow with a read intent for contents (prefer over `p.bash("find ...")` for static discovery) |
 | Persist generated output | `p.writeOutput(field, path)` or `p.writeInput(pathField, outputField)` |
 | String-keyed map | `s.record(value)`; keys are always `string` — do not wrap in `s.object`; use `s.record(s.int)` for count maps |
 | Numeric schema choice | `s.int` for counts/line numbers; `s.number` for measurements and ratios |

skills/rig/references/composition.md

@@ -140,6 +140,11 @@ When a task asks for runnable markdown:
 
 Rig starts with no addons. `maxTurns` is only the total budget; automatic parse/schema correction requires `repair()`:
 
+| Scenario | Addons | What happens |
+|----------|--------|--------------|
+| Invalid parsed or schema-validated output | `repair()` | Retries with the validation error |
+| The same failure needs stronger final guidance | `[steering(), repair()]` | Adds a last-chance warning to repair's final retry |
+
 ```ts
 import { agent, repair } from "rig";