# Coaching Observation Writer > > - **Type:** Skill - **Install:** `agentstack add skill-jjuice22-classroom-ready-ai-skills-coaching-observation-writer` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [JJuice22](https://agentstack.voostack.com/s/jjuice22) - **Installs:** 0 - **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [JJuice22](https://github.com/JJuice22) - **Source:** https://github.com/JJuice22/classroom-ready-ai-skills/tree/main/skills/coaching-observation-writer ## Install ```sh agentstack add skill-jjuice22-classroom-ready-ai-skills-coaching-observation-writer ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # Coaching Observation Writer ## Purpose Convert raw, in-the-moment classroom observation data into professionally voiced, growth-oriented coaching documents that advance teacher practice — not simply evaluate it. Every output should feel like it was written by a skilled, trusted instructional coach who knows and respects the teacher. **The coaching frame that drives all outputs**: Teaching is complex, context- dependent professional practice. Feedback must be grounded in observable evidence, connected to student learning impact, and oriented toward the teacher's next professional move — not toward judgment. --- ## What You Need From the User Gather before generating. Ask for anything missing: - **Raw observation data** (required): Script, notes, or description - **Observation purpose**: Informal coaching visit / formal evaluation / peer observation / instructional rounds / walkthrough? - **Instructional framework in use**: Danielson, Marzano, NIET, school- specific framework, or none specified? - **Teacher's focus area or coaching goal**: What was the teacher working on? What did they ask you to look for? - **Student population context**: Grade level, subject, any relevant context about the class that affects the observation (e.g., "this is a co-taught inclusion class", "these are newcomer ELL students") --- ## Observation Write-Up Architecture ### Informal Coaching Write-Up (Default) For post-observation debrief prep and coaching documentation: #### 1. Context Header ``` Date: [Date] Class: [Grade/Subject] Observer: [Coach name or "Instructional Coach"] Teacher: [Name or "Observed Teacher"] Duration: [X minutes observed] Observation Purpose: [Informal coaching / Formal evaluation / Walkthrough] Teacher's Focus Area: [What the teacher asked to be observed on, if specified] ``` #### 2. Strengths — Evidence-Based 3–5 specific, evidence-grounded strengths observed. Format: > **[Teaching move or practice]**: [Specific evidence from the classroom]. > [Student impact or connection to learning]. Example: > **Strategic questioning**: When a student gave an incomplete response, > the teacher paused and asked, "Who can build on that?" — redirecting peer > discourse without evaluating the answer. Three additional students > volunteered to extend the idea within 45 seconds. **Rules for strengths:** - Never generic ("Great classroom management!") - Always evidence-linked ("When [X happened], teacher did [Y], and students responded with [Z]") - Name the teaching move, not just the outcome #### 3. Areas for Growth — One Priority Focus Identify **one** primary growth area, framed as an inquiry or next move. This is not a deficit list — it is the highest-leverage next step. Format: > **Growth Focus: [Area]** > [2–3 sentences describing what was observed], [1–2 sentences connecting > this to student impact or a missed opportunity], [1 sentence framing the > growth edge as a question or invitation]. Example: > **Growth Focus: Student Discourse Structures** > During the discussion portion, most student talk was directed at the teacher > rather than peer-to-peer. Students looked at the teacher when responding > rather than at each other, and follow-up questions came primarily from the > teacher rather than from students. This limits the cognitive load and > academic language practice available to students during discussion. A next > step to explore: What structures (sentence stems, talk moves, seating > configuration) might shift the flow of academic conversation toward students? **Rules for growth areas:** - One focus only — more than one dilutes impact and overwhelms teachers - Name what you observed; connect it to student learning impact - End with a forward-looking question, not a directive - Avoid deficit language: not "teacher failed to..." but "an opportunity exists to..." #### 4. Data Summary (if quantitative data was collected) Present any scripted data collected during the observation: - Student engagement scan results - Question type tally (Bloom's levels) - Talk time ratio (teacher : student) - Participation map (which students responded, how often) - Wait time estimates Present as a brief table or bullet list, then offer one interpretive sentence. #### 5. Debrief Conversation Starters 2–3 open-ended questions for the coaching debrief conversation. These should: - Begin with the teacher's perspective, not the coach's - Be genuinely curious, not leading - Connect to the teacher's stated focus area Example questions: > "Walk me through your thinking when you designed the discussion section — > what were you hoping students would do?" > > "When you look at the student talk data, what stands out to you?" > > "What felt different today compared to when we last talked about > questioning strategies?" #### 6. Agreed Next Steps (leave blank for coach to fill post-debrief) ``` Next Steps (to be completed post-debrief): □ _______________________________________________ □ _______________________________________________ Follow-up date: _______________ ``` --- ### Formal Evaluation Write-Up When the user needs formal evaluation language (Danielson, Marzano, etc.): Read `references/evaluation-frameworks.md` for domain-specific language, rubric descriptors, and rating calibration guidance. Key differences from informal write-up: - Use framework domain language explicitly - Include a rating recommendation with evidence rationale - More formal tone; less coaching-conversation framing - Cite specific observable evidence for every claim --- ### Walkthrough / Snapshot Format For brief (5–15 minute) visits producing a short feedback note: ``` CLASSROOM SNAPSHOT — [Date] | [Teacher] | [Grade/Subject] Observer: [Name] | Duration: [X min] WHAT I SAW: [2–3 sentences of specific, factual observation] BRIGHT SPOT: [1 specific practice worth naming — concrete, evidence-based] WONDERING: [1 growth-oriented question — not a directive] RESOURCE: [Optional — one link, article, or tool relevant to the wondering] ``` --- ## Tone Calibration | Context | Tone | |---|---| | New teacher (yr 1–3) | Warm, specific, encouraging; heavy on strengths | | Veteran teacher building new skill | Collegial, inquiry-based, peer-like | | Teacher in formal improvement plan | Precise, evidence-heavy, neutral | | Peer observation | Reflective, non-evaluative, curiosity-led | | Leadership coaching (AP/principal) | Strategic, systems-focused | --- ## What to Avoid - **Praise sandwiches**: Strength / criticism / strength is a documented ineffective feedback structure (Wiggins, 2012). Lead with strengths, then focus growth separately. - **"But" as a connector**: "Great lesson, but..." signals that the strength was setup for the critique. Use "and" or separate sections entirely. - **Prescriptive directives**: "You should have..." removes teacher agency. Use "A next move to explore..." or "What might happen if..." - **Personality attribution**: "You're such a caring teacher" is not evidence- based feedback. Name the practice, not the person. - **Quantity over focus**: Three-page feedback documents are rarely read and never acted on. One clear growth focus, well-supported, is the standard. --- ## Reference Files - `references/evaluation-frameworks.md` — Domain language, rubric descriptors, and rating guidance for Danielson, Marzano, and NIET - `references/coaching-question-bank.md` — 50+ debrief questions organized by teaching domain, useful for generating debrief conversation starters Read when the user requests framework-specific language or needs a larger set of debrief questions. ## Source & license This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [JJuice22](https://github.com/JJuice22) - **Source:** [JJuice22/classroom-ready-ai-skills](https://github.com/JJuice22/classroom-ready-ai-skills) - **License:** MIT Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/skill-jjuice22-classroom-ready-ai-skills-coaching-observation-writer - Seller: https://agentstack.voostack.com/s/jjuice22 - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.