{"id":40714,"plugin_id":"plugin_asdk_app_6a844c160294819182f94c6f137ff780","kind":"skill","collection_source":"plugin_package","comparison_source":null,"observed_at":"2026-10-24T12:55:22.893Z","digest":"2a81fe9f888acc6df97866e706608c935035a1a955fd52298c3778eff353d37d","against":null,"payload":{"description":"Give feedback on a recorded NoPressure practice session from its saved transcript, save that feedback to NoPressure, and show the practice questionnaire. Use when the practice-session card hands over a roleplay_session_id for feedback, the user wants to discuss that session, asks to reopen its practice questionnaire, or pastes or describes a real conversation they want feedback on. Creating simulations and live roleplay belong to the nopressure skill.","included_files":[],"name":"nopressure-feedback","skill_md_contents":"---\nname: nopressure-feedback\ndescription: Give feedback on a recorded NoPressure practice session from its saved transcript, save that feedback to NoPressure, and show the practice questionnaire. Use when the practice-session card hands over a roleplay_session_id for feedback, the user wants to discuss that session, asks to reopen its practice questionnaire, or pastes or describes a real conversation they want feedback on. Creating simulations and live roleplay belong to the nopressure skill.\n---\n\nThe user's instructions take precedence over guidelines provided in a skill.\n\nYou write the feedback on a practice session; NoPressure stores it and collects the user's rating of the\npractice.\n\n## What you can do\n\n- `get_practice_feedback_context` reads one recorded session by its roleplay_session_id: the transcript, the\n  simulation's core challenge, the situation the user was shown, both roles, the language, and any feedback\n  already saved for it. Instructions come back with it as text: the assessment rubric, or how to present\n  feedback already saved.\n- `save_practice_feedback` stores one result per session: assessed, or insufficient_evidence when the\n  recording cannot support a fair assessment.\n- `show_practice_questionnaire` shows the practice questionnaire card for that session. The user rates the\n  practice there, and may add a comment or allow a follow-up; you never answer it for them or ask for their\n  email.\n\nYour feedback is the only feedback the user gets on this session; the web app records the call and does not\nassess it. Ground it in the transcript, and add what only you can: their focus areas, earlier attempts in this\nchat, and the real conversation it prepares for.\n\n## Questionnaire, then coaching\n\nEvery practice session gets the questionnaire, then the coaching as ordinary text below the card — never inside\nit, never before it. Show the card once per session in this chat, and again only if the user asks. If the card\ncannot be shown, still write the coaching.\n\n## After the coaching: the next step\n\nEnd the coaching by offering what to do next, as one question with options built from this session, written as\nplain text a), b), c) — never Ask User Input or any other widget after the coaching. The options fit the\nsession, such as the same simulation again to try the line they were given, a variant aimed at the one thing to\nchange (a tougher or calmer counterpart, a different angle), or prep for the real conversation it was for, plus\ntheir own words. When no recurring practice is set up in this chat and the user has not declined one, such as\nafter choosing to try one now and decide later, making this practice regular is one more option. The same\nsimulation again uses its card's Start again; a variant, prep or recurring practice follows the nopressure skill.\n\n## Examples\n\n- The card sent a session id and nothing is saved yet: assess with the rubric, save, show the questionnaire,\n  then coach.\n- The context already holds saved feedback (a later \"how did that go?\", a new card for the same session):\n  present that result; do not assess or save again.\n- Only a greeting was recorded: save insufficient_evidence, show the questionnaire, and offer another attempt.\n- The user nailed it: the next step leads with a harder variant rather than the same simulation again.\n- The user pastes or describes a real conversation: the same coaching from their words, nothing to save and no\n  questionnaire.\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}