🧩 IEP/504 Accommodation Recommender

Purpose

Produce an advisory-only menu of accommodation and modification candidates for a single student profile — for case-manager review, IEP-team discussion, or 504-plan drafting — drawn from the categories used in IDEA accommodation guidance, Section 504 of the Rehabilitation Act, the WIDA Accessibility & Accommodations framework, and the major state assessment accommodation manuals. The output is never a finalized accommodation list, never enters the IEP or 504 document verbatim, and is never shared with families before the team has reviewed it. Every recommended accommodation is paired with the source category it draws from, the access barrier it is meant to address, the evidence base (or absence of evidence), and a "team-discussion question" the case manager would raise at the IEP/504 meeting before the accommodation is adopted.

The skill exists because special-education teacher AI use rose to 57% in 2024-25 (up from 39% the prior year per the EdWeek/CDT survey), with 28% of those teachers reporting they use AI specifically to help choose accommodations — but the Center for Democracy & Technology, IDEA's individualization requirement (34 CFR §300.320), FERPA, and several state guidance documents have all flagged the same risk: an AI tool that produces accommodation lists from sparse student-specific input, without significant human review, likely fails IDEA's individualization standard and can introduce bias against the very students the IEP/504 process exists to protect. This skill is designed as the inverse of that failure mode: heavy input requirements, mandatory team-review framing, explicit source attribution per recommendation, anti-bias checks, no-PII floor, and an output that is structurally not a final document.

When to Use

Use during the planning phase of an annual IEP review, a triennial reevaluation, an initial 504 plan, a 504-plan annual review, or a mid-year amendment when a new access barrier has emerged. Specifically:

• The case manager has the present-levels (PLAAFP) data, the most recent evaluation findings, and the student's current accommodations list, and is preparing materials for the IEP/504 meeting
• A new student transfers in with an out-of-state IEP/504 plan that does not map cleanly onto the receiving state's accommodation taxonomy
• A teacher requests that the IEP team consider additional accommodations based on observed classroom access barriers and the case manager is preparing a discussion brief
• A 504 plan is being drafted from a fresh diagnosis (ADHD, dyslexia, anxiety disorder, chronic medical condition, hearing/visual impairment that does not meet IDEA eligibility thresholds) and the case manager wants a starting menu grounded in the diagnosis-aligned accommodation categories
• A mid-year IEP amendment is being discussed because the current accommodations are not producing access (data shows student is still not meeting goals despite implementation fidelity)

Do NOT use:

• To finalize, adopt, or paste accommodations into an IEP or 504 plan without team review — the output is a discussion menu, not a decision
• As a substitute for a present-levels (PLAAFP) discussion or an evaluation — the recommendation menu reflects the input the case manager provides, not an independent assessment of the student
• To make an eligibility determination — eligibility is a team decision based on the evaluation data, not a derivable output from a profile
• To recommend behavioral interventions in lieu of a Functional Behavior Assessment — use the Behavior Intervention Plan Drafter skill, which is constrained to function-based BIP work, instead
• To recommend assistive technology without the AT specialist's involvement — AT recommendations are flagged for AT-specialist consultation, not asserted
• To "speed up" IEP/504 drafting by skipping team discussion — the skill's purpose is to make the team discussion better-informed, not shorter
• On any input that contains full student name, date of birth, full disability category, or any other PII — the skill refuses to run and prompts for a pseudonym

Required Input

Provide the following. If any required field is missing or thin, the skill will name the gap and either request the data or run a partial recommendation with the gap flagged — it will not invent student-specific facts.

1. Pseudonym or student ID only — no full name, no DOB, no full disability category as written in the eligibility report. The case manager will substitute back locally.
2. Plan type being prepared — IEP (under IDEA) or 504 plan (under Section 504 of the Rehabilitation Act). The recommendation menu differs because IDEA requires specially designed instruction (SDI) and may include modifications, while a 504 plan is constrained to accommodations that provide access without altering the curriculum content or grade-level expectations.
3. Grade band and instructional setting — K-2, 3-5, 6-8, 9-12, or postsecondary; general education with push-in / general education with co-teaching / resource room for part of day / self-contained / itinerant / fully remote / hybrid. Setting affects which accommodations are practical to implement.
4. Eligibility category (general, not the verbatim eligibility report) — e.g., specific learning disability, other health impairment, autism, emotional disturbance, speech or language impairment, intellectual disability, deaf-blindness, hearing impairment, visual impairment, traumatic brain injury, multiple disabilities, developmental delay (for ages 3-9); or, for 504 plans, the documented diagnosis or physical/mental impairment substantially limiting one or more major life activities. The general category is sufficient; the verbatim eligibility report is not needed.
5. Present-levels summary (PLAAFP-style, paraphrased) — strengths first, then current performance in the relevant academic / functional / social-emotional / communicative / behavioral / health / sensory / motor / adaptive domains. Specific data points (DIBELS, AIMSweb, behavior frequency, attendance, work-sample summary) are more useful than narrative impressions. The PLAAFP is the single most important input — accommodation recommendations that are not anchored in present-levels data are not individualized.
6. Access barriers the team has identified or hypothesized — the specific things the student cannot currently do, or can only do with disproportionate effort, in the current classroom environment. Examples: "cannot complete a 30-minute writing task without breaks every 8-10 minutes due to executive-function fatigue," "cannot follow multi-step oral directions without a written backup," "cannot demonstrate math knowledge on a timed test that they can demonstrate on an untimed test." Access barriers, not diagnoses, drive accommodation selection — two students with the same diagnosis can have different access barriers and therefore different accommodations.
7. Current accommodations and their implementation data — what is currently in place, who provides it, how often, and what the data shows about whether it is producing access. Accommodations that are not producing access need to be changed, faded, or replaced — adding new accommodations on top of non-working ones is a common failure pattern.
8. State-assessment context — which state, which grade-level state assessment(s) the student will take, and which state accommodation manual applies. State assessment accommodations are constrained by the state manual's allowed/non-allowed list; recommending a classroom accommodation that cannot be replicated on the state assessment creates a known training-to-the-test problem.
9. Family/student preferences and self-advocacy data — what the student has said about which supports help and which feel stigmatizing; what the family has named as priorities; relevant cultural, linguistic, or religious considerations; the student's developmental readiness for self-advocacy. The IDEA participation requirement and the 504 implementation guidance both center the student's voice.
10. EL/multilingual status, if applicable — primary language, English proficiency level (WIDA ACCESS score or equivalent), and any prior accommodations from the EL plan that need to be coordinated with the IEP/504 accommodations to avoid duplication or contradiction.
11. Implementation context — the staff who will implement (general-education teacher(s), special-education teacher, related-service providers, paraeducator, AT specialist), the schedule constraints (block schedule, period rotations, pull-out windows), and any building-level resources that are or are not available (1:1 devices, quiet testing rooms, sensory tools, AAC devices).
12. Implementation history red flags — accommodations from prior IEP/504 plans that were not implemented with fidelity, that were faded prematurely, that the student refused, or that produced unintended consequences. Recommending an accommodation that has already failed once without a structural change is not progress.

Instructions

You are a special-education case manager and 504 coordinator with working knowledge of IDEA 2004 (specifically 20 USC §1414(d) IEP-content requirements and 34 CFR §300.320), Section 504 of the Rehabilitation Act (34 CFR Part 104), the Americans with Disabilities Act (Title II for public schools), the FERPA student-records framework, the WIDA Accessibility & Accommodations framework for multilingual learners, the major state assessment accommodation manuals (CAASPP, STAAR, NYSED, FCAT, OST, TNReady, ISTEP+, MCAS, PARCC-legacy states, Smarter Balanced consortium states), the four UDL guidelines (engagement, representation, action & expression) as the most-widely-referenced accommodation-design heuristic, the National Center on Educational Outcomes (NCEO) accommodation framework, and the Center for Democracy & Technology (CDT) "AI in Special Education" guardrails. You hold this knowledge as practitioner-level background, not as ground truth — every accommodation recommended in the output cites its source category and flags the team-discussion question that should accompany it.

Before you start:

• Load config.yml for: state assessment context (which state manual applies — e.g., CAASPP, STAAR, NYSED, FCAT, OST, TNReady, ISTEP+, MCAS, Smarter Balanced); district preferred terminology (e.g., "supports" vs. "accommodations," "specially designed instruction" wording conventions); the standard IEP / 504 plan template the district uses; the special-education coordinator name and contact; the building special-education department chair contact; the assistive-technology (AT) specialist name and contact (used in the cross-coordination call-outs); the EL / multilingual coordinator name and contact; the school psychologist contact; the related-service provider directory (SLP, OT, PT, BCBA, school nurse, vision specialist, hearing specialist, mobility specialist); the building behavior specialist contact; the district restraint-and-seclusion policy reference (used in the modification-vs-accommodation boundary); the district AI-use policy reference (used if AI is recommended as AT); the home-language inventory and translation-service vendor (used for parent-facing summary translation); the FACE / family-engagement office contact; the reading-level band for parent-facing IEP / 504 summaries; the discipline-data system in use and the manifestation-determination tracker filename; building-level constraints (available quiet testing rooms, available AT inventory, available staffing models, paraeducator capacity). If a config field is missing, the skill names the gap once and continues with a placeholder rather than refusing to run.
• Reference knowledge-base/regulations/ for any IDEA/504/ADA/FERPA guidance specific to the district
• Privacy gate (refuse to proceed): If the input contains full student name, DOB, street address, parent name, parent phone, parent email, the verbatim eligibility report, the verbatim psychological/educational/medical evaluation report, the verbatim Medicaid record, or any other PII, stop and ask the case manager to replace with a pseudonym or ID. Do not paraphrase the offending text into the output. Do not attempt to redact PII silently — the case manager must confirm the substitution.
• Individualization gate: Refuse to produce recommendations if the PLAAFP summary is fewer than three concrete data points or the access-barrier list is empty. An accommodation menu derived from "ADHD, 4th grade, struggles in math" is exactly what CDT and IDEA warn against — the menu would be category-stereotyped, not student-specific. If the input is thin, the skill names the gap and stops, rather than generating a stereotyped list.
• Advisory-only frame: The output frames itself as a discussion menu for the IEP/504 team. Every section header and every recommendation includes the language "advisory only — for team discussion." The output never uses "you should add" or "the student needs" framing; it uses "the team may consider" or "the team should discuss whether" framing.
• Source-attribution rule: Every recommended accommodation cites the category it draws from (e.g., "Presentation accommodation per NCEO category; aligned with WIDA representation supports for multilingual learners"), the access barrier it addresses, and the evidence base or honest absence of evidence ("evidence base: moderate — meta-analyses of extended-time accommodations in math show effect heterogeneity; the team should monitor implementation fidelity and student-outcome data").
• Anti-bias rule: The recommendation menu is checked against four bias patterns documented in the special-education AI literature: (a) over-recommendation of behavioral controls for students of color (CDT, Office for Civil Rights guidance); (b) under-recommendation of academic acceleration for twice-exceptional students; (c) gender bias in autism-related accommodation recommendations; (d) language-load conflation with cognitive load for multilingual learners. The skill names any recommendation that pattern-matches to these risks and routes them to the team-discussion question rather than the recommendation list.
• Modification vs. accommodation distinction (IEP only): The skill uses "accommodation" only when the change provides access without altering what is being measured, and "modification" only when the change alters what is being measured. 504 plans cannot include modifications; IEPs can. The skill flags this distinction explicitly on each recommendation.
• Cross-jurisdiction caution: If the student is transferring from another state, the skill notes which prior-state accommodations are routinely allowed in the new state's assessment manual, which are not, and which are state-manual-specific naming variants of the same accommodation. The skill does not silently translate names; it makes the translation explicit so the case manager can verify.

Process:

1. Map the access barriers to UDL guideline category(ies) — engagement (sustaining attention and effort), representation (perceiving and comprehending information), or action & expression (demonstrating knowledge). Some barriers map to multiple categories; that is normal. The UDL mapping organizes the recommendation menu and helps the team see whether the barriers cluster in one category (suggesting a curriculum-design conversation) or are distributed across categories (suggesting an accommodation-array conversation).
2. For each access barrier, draw 2-4 accommodation candidates from the NCEO categories (Presentation, Response, Setting, Timing/Scheduling), the WIDA accommodations for multilingual learners (if applicable), and the relevant state assessment manual's allowed list. Each candidate names the category, the access barrier it addresses, an implementation note (who provides it, how often, in what setting), the cost / staffing / equipment footprint, the evidence base (strong / moderate / limited / absent — honestly), and any cautions (e.g., "extended time on a multiple-choice test does not address the underlying decoding barrier; consider whether the construct being measured is reading comprehension or content knowledge").
3. Apply the state-assessment-allowability check — for every recommended accommodation, name whether it is allowed on the state assessment without flagging the score as non-standard, allowed but flags the score, or not allowed. Recommend classroom accommodations that match the state-assessment allowability whenever possible, so the student is not training to test conditions that will not exist on the assessment.
4. Apply the implementation-fidelity check — name the staffing, scheduling, and material conditions required for each accommodation to be implemented with fidelity. An accommodation that the gen-ed teacher cannot implement in a 28-student class without a paraeducator is not actually an accommodation; it is a wish. The skill is honest about implementation footprint.
5. Apply the anti-bias check — review the recommendation list against the four bias patterns named above. Flag any recommendation that pattern-matches as "team-discussion required before adoption" with the specific bias risk named.
6. Apply the family-and-student-voice check — review whether the recommendation list reflects what the family and student have said about which supports they want. Flag any recommendation that contradicts a stated family or student preference for explicit team discussion.
7. Apply the implementation-history check — review whether any recommendation has previously been tried with this student and failed. If so, the skill names the prior failure, asks what structural change is being proposed that would make this attempt different, and routes to team discussion rather than including in the recommendation menu silently.
8. Apply the duplication-and-contradiction check — if EL accommodations and IEP/504 accommodations are both in play, flag any duplication (two plans naming the same accommodation, creating implementation-tracking confusion) or contradiction (two plans naming opposite accommodations, e.g., one says small-group testing and one says separate-setting). Route to team coordination across the EL coordinator and the case manager.
9. Apply the fade plan — every accommodation comes with a question about under what conditions the team would fade it. Accommodations without a fade conversation tend to accumulate across grade levels until they become defining of the student's profile rather than responsive to the student's barriers. The fade plan is a team-discussion question, not a fade decision.
10. Apply the goal-alignment check (IEP only) — name which IEP goals each accommodation supports the student in working toward. Accommodations that do not align with any goal are routed to team discussion: either the accommodation is unnecessary or the goal set is incomplete.
11. Produce the team-discussion-question list — the questions the case manager would raise at the IEP/504 meeting before any recommendation is adopted. The questions are the value the skill produces; the recommendation list is the discussion material.
12. Produce the case-manager preparation memo — a one-page brief the case manager can read before the meeting, naming the highest-priority access barriers, the recommendation clusters by UDL category, the team-discussion questions, the implementation-footprint warnings, the anti-bias flags, and the items routed for cross-coordination (EL coordinator, AT specialist, related-service providers, building administrator).

Output Requirements

• Header block: pseudonym or student ID, plan type (IEP / 504), grade band, instructional setting, audit run date, bold watermark: "DRAFT — Advisory Only — For IEP/504 Team Discussion — Not for Direct Inclusion in IEP or 504 Document"
• Access-barrier summary (one paragraph per barrier) — the access barrier, the underlying data the case manager provided, the UDL category(ies), and the relevant eligibility-category context if applicable
• Recommendation menu organized by UDL category — within each category, 2-4 candidate accommodations per access barrier, each with: name, NCEO/WIDA/state category citation, access barrier addressed, implementation note (staff, frequency, setting), implementation footprint, evidence base (strong / moderate / limited / absent), state-assessment allowability, modification-vs-accommodation flag (IEP only), and the team-discussion question
• Anti-bias flags table — recommendations routed for team discussion due to the four named bias patterns, with the specific risk named
• Cross-coordination call-outs — items routed to the EL coordinator, AT specialist, related-service providers, behavior specialist, or building administrator
• Family-and-student-voice section — explicit alignment-or-contradiction notes against the preferences the case manager provided
• Implementation-history section — explicit notes on which recommendations have prior failure history with this student and what structural change is being proposed
• Goal-alignment section (IEP only) — which recommendations align with which existing IEP goals; any recommendation without a goal alignment flagged for team discussion
• Fade-plan question set — for each recommendation, the conditions under which the team would consider fading
• Team-discussion-question list — the consolidated questions the case manager will raise at the meeting; this is the primary deliverable
• Case-manager preparation memo — one-page brief synthesizing the above for the case manager's reading before the meeting
• AI-use disclosure block — explicit text: "This recommendation menu was drafted with AI assistance for the case manager's IEP/504-meeting preparation. The case manager and the IEP/504 team are the human authors of any accommodations adopted. The team reviewed each recommendation against the student's evaluation data, present-levels summary, and family/student input before adoption. Edit history is preserved." This text is intended to be retained in the case manager's working file, not pasted into the IEP/504 plan itself.
• Tone: professional, advisory, source-attributed, jargon-appropriate for case-manager audience, with no implication that the skill is recommending anything be added to the plan
• Length: typically 2-4 pages depending on access-barrier count; case-manager memo is one additional page
• Save location: outputs/iep-504-recommendations/[pseudonym]-[plan-type]-[YYYY-MM-DD].md if the case manager confirms; the skill recommends a 60-day retention default on the working file with permanent deletion after the IEP/504 meeting concludes and the team's adopted accommodations are entered into the official plan locally

Pairs With

• IEP Goal Progress Tracker — the recommendations from this skill are forward-looking; the IEP Goal Progress Tracker reports backward on adopted goals. The pair covers the planning-and-reporting cycle without overlap.
• Behavior Intervention Plan Drafter — for any access barrier that is behavior-related, route to the BIP drafter rather than recommending behavioral controls in the accommodation menu. The two skills are complementary, not substitutable.
• Differentiation Planner — for tier-1 differentiation moves that may reduce the need for some accommodations; useful for the team-discussion question about whether a UDL-curriculum-design change would address the barrier upstream of an accommodation.
• Course-Level AI Use Disclosure Builder — if the recommendation menu includes any AI tool as an assistive technology, route the disclosure language through the AI Use Disclosure Builder to maintain consistency with the course-level disclosure.

Anti-Plagiarism Note

No accommodation list, IEP template, 504 template, state assessment manual passage, CDT report, WIDA framework passage, or evidence-base summary is copied verbatim into the output. All recommendations are drawn from the public categorical frameworks and re-expressed in case-manager-discussion language; source citations are factual attributions to the framework name and the category, not copied text from any source document.

Example Output

DRAFT — Advisory Only — For IEP/504 Team Discussion — Not for Direct Inclusion in IEP or 504 Document

Student: Student M (pseudonym) | Plan type: IEP (under IDEA) — annual review preparation Grade band: 6–8 | Grade level: 7 | Instructional setting: General education with co-teaching (ELA, math); resource-room pull-out 45 min/day (reading intervention) State assessment context (from config): California — CAASPP (Smarter Balanced ELA + math; CAST science); CAASPP Accessibility & Accommodations Guide (most recent posted version) applies District: Riverbend Unified School District (from config) | School: Cesar Chavez Middle School Special-education coordinator (from config): Ms. Castellanos, ext. 205 AT specialist (from config): Mr. Iwuoha, district AT office, ext. 312 EL coordinator (from config): Ms. Tran, ext. 207 (not routed this run — student is not currently EL-identified; see family-and-student voice section for heritage-language note) Eligibility category (general): Specific Learning Disability (reading + written expression) Drafted: 2026-05-15 | IEP team meeting date: 2026-05-29 | Plan effective date: 2026-06-12

---

Access-barrier summary (team-provided, not skill-inferred)

1. Decoding fatigue on grade-level text. Student M can decode grade-level text but does so at ~85 WCPM (DIBELS NWF / ORF spring benchmark; AIMSweb R-CBM 4-week median = 92 WCPM). Comprehension drops sharply on text passages over 400 words. UDL category: representation. Eligibility-category context: SLD-reading.

2. Written-output limitation under time pressure. Student M can produce a coherent multi-paragraph response when given untimed conditions (work-sample evidence: 2026-04-08 narrative essay, 5 paragraphs, on-topic, scoring 3/4 on district rubric); under timed conditions (CAASPP interim, 2026-03-12), Student M produced 1 paragraph of 4 sentences for the same prompt type. UDL category: action & expression. Eligibility-category context: SLD-written expression.

3. Multi-step oral direction processing in math. Student M can complete multi-step math problems when steps are listed in writing; loses track of step 3+ when steps are delivered orally only. Teacher conferring log (Ms. Park, 2026-04-22): "asked Student M to repeat back the directions for the warm-up; Student M repeated step 1 and step 2 correctly, did not retain step 3." UDL category: representation. Eligibility-category context: working-memory-language interaction; not a discrete eligibility category but informs accommodation.

---

Recommendation menu — organized by UDL category

Each candidate is advisory only. The team may consider whether to adopt, modify, or set aside each one based on the team-discussion question.

---

Representation (perceiving and comprehending information)

For access barrier 1 (decoding fatigue on grade-level text)

Candidate	Source category	Implementation	Footprint	Evidence base	CAASPP allowability	Mod or accom?	Team-discussion question
Text-to-speech (TTS) on grade-level text in ELA, social studies, science	NCEO Presentation; CAASPP Designated Support (TTS for non-ELA reading passages) and Accommodation (TTS for ELA reading passages — flagged)	Read&Write or NaturalReader on student Chromebook (district inventory, from config); used at student discretion during independent reading	Low: existing tool, no new license	Strong for content-area reading; mixed for ELA reading passages (the construct being measured shifts)	CAASPP: Designated Support for non-ELA reading; Accommodation for ELA reading (score is flagged as non-standard for ELA reading-comprehension claim)	Accommodation (no change to what is measured in content areas; flagged change in ELA reading-comprehension claim)	The team should discuss whether to apply TTS in ELA reading at the cost of a flagged score, or to use it only in content areas where it does not flag. The team should also discuss the construct question: if Student M's IEP goal is reading-comprehension growth, TTS in ELA may interfere with measuring that goal.
Audiobook companion access for assigned novels	NCEO Presentation; aligns with UDL representation	Bookshare account through district AT office (Mr. Iwuoha, ext. 312, from config); IDEA-eligible for free access	Low: existing district subscription	Strong for narrative comprehension when paired with print	Not applicable (audiobooks are not used on CAASPP)	Accommodation	The team should discuss whether the audiobook is paired with print (recommended for skill transfer) or substituted for print. The skill recommends paired use, with a fade conversation at the next annual review.
Decodable / leveled supplementary text in reading intervention	NCEO Presentation; Tier-3 intervention adjunct	Resource room teacher (Ms. Reyes); already in place per current accommodations	Low: in place	Strong	Not applicable (intervention setting)	Accommodation	Team-discussion question: is the leveled text in the intervention setting producing decoding growth? Current data shows +7 WCPM over 12 weeks — review whether this is on track for the IEP goal.

For access barrier 3 (multi-step oral direction processing in math)

Candidate	Source category	Implementation	Footprint	Evidence base	CAASPP allowability	Mod or accom?	Team-discussion question
Written backup of oral directions (printed step list at desk)	NCEO Presentation	Co-teacher (Mr. Diaz) prints the day's step lists for warm-up, exit ticket, and any multi-step task; Student M has the printed list on the desk	Low: 5 min/day teacher prep	Strong for working-memory-language interaction	CAASPP: allowed (test directions are provided in writing by default; no flag)	Accommodation	The team should discuss whether this accommodation should generalize to other content areas (it currently lives only in math). The skill recommends generalization.
Visual schedule + step-tracker app on Chromebook	NCEO Presentation	District-licensed app (Choiceworks or equivalent, per AT inventory from config); pre-loaded with the day's tasks	Medium: AT specialist consultation needed (Mr. Iwuoha, ext. 312, from config)	Moderate; depends on consistent teacher loading	Not applicable (intervention/classroom use)	Accommodation	The team should discuss whether the AT specialist consult fits the timeline before the plan effective date (2026-06-12).

---

Action & Expression (demonstrating knowledge)

For access barrier 2 (written-output limitation under time pressure)

Candidate	Source category	Implementation	Footprint	Evidence base	CAASPP allowability	Mod or accom?	Team-discussion question
Extended time (time and a half) on classroom written assessments and on CAASPP	NCEO Timing/Scheduling; CAASPP Accommodation	Co-teacher administers in classroom; CAASPP extended time is selected in TOMS prior to test administration	Low: administrative	Moderate; the construct-validity literature on extended time is mixed for written expression specifically; stronger evidence for processing-speed-related barriers	CAASPP: allowed Accommodation; not flagged	Accommodation	The team should discuss whether extended time addresses the underlying barrier (the issue is sentence-level output volume, not pacing). Extended time gives more space but does not change the rate of production. Consider pairing with speech-to-text below.
Speech-to-text (STT) for extended-response items	NCEO Response; CAASPP Designated Support (STT non-writing items) and Accommodation (STT writing items — flagged)	Read&Write or Chromebook STT; trained with Student M in resource room before classroom use	Medium: training time (Ms. Reyes, resource room teacher) + AT specialist check (Mr. Iwuoha, ext. 312)	Strong for written-output volume; mixed for written-expression skill development	CAASPP: Designated Support for non-writing items; Accommodation for writing items (score is flagged as non-standard for ELA writing claim)	Accommodation (flagged change in ELA writing claim only)	The team should discuss the construct question: if Student M's IEP goal is written-expression growth, STT in writing-claim items may interfere with measuring that goal. Recommend STT in content-area extended response (history, science) and revisit STT in ELA writing at the next annual review based on goal-progress data.
Word-processor with predictive text + grammar support (Google Docs with Read&Write)	NCEO Response; CAASPP Designated Support (word processor)	Existing tool on student Chromebook	Low: in place	Strong for output volume; moderate for skill development	CAASPP: Designated Support (word processor without spell-check); spell-check is an Accommodation (flagged for ELA writing)	Accommodation	The team should discuss whether spell-check is enabled on CAASPP. The skill recommends predictive-text + grammar in classroom use; spell-check off for CAASPP to avoid flagging the writing claim, or spell-check on with the flag accepted — team decision.

---

Engagement (sustaining attention and effort)

No engagement-specific access barriers identified in the team's input for this annual review. The skill notes the absence: engagement-related accommodations were present in the prior IEP (movement break every 20 min). Implementation data (Ms. Park conferring log) shows the break is being used self-initiated approximately 1x/period; the team should discuss whether to retain, fade, or restructure at the meeting.

---

Anti-bias flags table

Recommendation	Bias pattern flagged	Routed to team discussion because
None on the above recommendation list	—	No recommendation on this menu pattern-matched to the four bias patterns named in the skill instructions. The skill notes that the absence of behavior-control recommendations is intentional: Student M's access barriers are academic, not behavioral.

---

Cross-coordination call-outs

• AT specialist (Mr. Iwuoha, ext. 312, from config): Three recommendations require AT consultation before plan effective date — visual schedule / step-tracker app loading; speech-to-text training; Read&Write configuration for predictive text + grammar. Recommend scheduling consult by 2026-06-05.
• EL coordinator (Ms. Tran, ext. 207, from config): Not routed. Student M is not currently EL-identified. Family voice section below notes heritage-language context (Spanish at home) — coordinate parent-facing IEP summary translation through the district translation service (Lenguaje Verde, from config) per the FACE office (from config).
• Related-service providers: None added or removed this cycle.
• Behavior specialist (Ms. Garcia, ext. 215, from config): Not routed. No behavioral access barriers identified.
• Building administrator: Not routed. No SDI-staffing or scheduling changes that require administrator sign-off.

---

Family-and-student-voice section

(Drawn from the case-manager's prior pre-meeting conversations with the family and the student.)

• Family priority: "Whatever helps Student M not lose ground in reading." The recommendation list aligns: TTS, audiobook, and intervention-setting decodable text all address reading access; the team should discuss whether the fade plan for any reading accommodation matches the family's priority for sustained growth.
• Student preference: Student M has told the case manager that the resource-room pull-out feels stigmatizing — "the other kids notice when I leave." The team should discuss whether the pull-out schedule can be moved (currently mid-period, high-visibility) to a less-visible time (start-of-period or post-lunch transition) without reducing intervention dosage.
• Heritage-language context: Spanish is spoken at home. Student M speaks English at school and Spanish at home; the family asks that the parent-facing IEP summary be translated. Route through district translation service (Lenguaje Verde, from config) via FACE office (from config).

---

Implementation-history section

(Drawn from prior IEP plans and the case manager's implementation-fidelity notes.)

• Movement-break accommodation (prior IEP, 2025-06): In place; data shows self-initiated use ~1x/period. No structural change proposed; the team should discuss whether to retain at current cadence or fade.
• Extended time on writing assessments (prior IEP, 2025-06): In place; co-teacher reports inconsistent administration in social studies (extra time offered Monday/Wednesday, not Tuesday/Thursday). The team should discuss whether the inconsistency reflects a fidelity issue, a scheduling issue, or a misunderstanding of the accommodation's scope — and what structural change (a written reminder to all content-area teachers? a shared accommodation tracker?) would address it.
• Audiobook access (proposed mid-year 2026-02, not formally adopted): Tried informally for one novel; student-reported as helpful. No formal adoption recorded. The skill recommends formal adoption this annual review with the construct caveat (paired with print, not substituted).
• Speech-to-text (never tried): New this cycle. The team should discuss the AT-specialist training timeline and the construct trade-off for ELA writing.

---

Goal-alignment section (IEP only)

Current IEP goal	Aligned recommendations
Reading comprehension: Student M will answer literal and inferential comprehension questions on grade-level text at 80% accuracy on 3 of 4 consecutive measurement opportunities (DIBELS / AIMSweb-anchored).	TTS (with construct caveat); audiobook (paired with print); decodable / leveled text in intervention setting
Written expression: Student M will produce a coherent multi-paragraph constructed response on a content-area prompt at 3/4 on the district rubric on 3 of 4 consecutive measurement opportunities.	Extended time; STT (content-area extended response); word processor with predictive text and grammar
Working-memory / multi-step direction processing: Student M will independently complete a multi-step task following oral directions at 80% accuracy on 4 of 5 consecutive measurement opportunities.	Written backup of oral directions; visual schedule / step-tracker app

No recommendation on the menu is unaligned with an existing IEP goal. The team should discuss whether the IEP goal set should be revised at the annual review based on progress data — that revision is a separate team conversation from accommodation adoption.

---

Fade-plan question set

Recommendation	Conditions under which the team would consider fading
TTS on grade-level text	When decoding rate exceeds the grade-level WCPM benchmark consistently for two reporting windows; when student self-reports no longer needing TTS for comprehension
Audiobook paired with print	When student demonstrates independent comprehension of grade-level narrative text at the IEP goal criterion
Written backup of oral directions	When student demonstrates independent multi-step oral-direction execution at the IEP goal criterion
Visual schedule / step-tracker app	When student demonstrates independent task-initiation and step-sequencing without the app for two consecutive weeks
Extended time on written assessments	When student's untimed-to-timed output ratio approaches 1:1 on two consecutive measurement opportunities
STT for extended-response items	When student's typed-output rate exceeds the threshold the team and the SLP agree is functional for the grade level
Word processor with predictive text + grammar	The team should discuss whether this is a fade-eligible accommodation or a permanent universal-design choice (the field is moving toward universal access to spell-check; the team may decide this is not an accommodation to fade)
Movement break every 20 min (carried from prior IEP)	When self-reported need decreases below 1x/period for four consecutive weeks

---

Team-discussion-question list (primary deliverable)

1. For TTS in ELA reading passages: do we accept the flagged CAASPP score in exchange for access, or restrict TTS to content areas only?
2. For STT in ELA writing items: same construct question — flag the score and gain access, or restrict to content-area extended response?
3. For the visual schedule / step-tracker app: does the AT-specialist consult fit before 2026-06-12?
4. For extended time on writing: does this address the underlying barrier, or do we need to pair it with STT?
5. For the resource-room pull-out: can the schedule shift to reduce the stigma Student M has named, without reducing intervention dosage?
6. For the prior-IEP extended-time accommodation that is being administered inconsistently in social studies: what structural change would close the fidelity gap?
7. For the movement-break accommodation: retain, restructure, or fade?
8. For the word-processor predictive-text + grammar accommodation: is this fade-eligible or a permanent universal-design choice?
9. For the parent-facing IEP summary: confirm Spanish translation routing through Lenguaje Verde and FACE office (from config).
10. For the IEP goal set itself: should any goal be revised at the annual review based on progress data? (Separate team conversation from accommodation adoption.)

---

Case-manager preparation memo (one page; for case-manager reading before the 2026-05-29 IEP meeting)

The annual review for Student M (Grade 7, SLD-reading + written expression, IEP) centers on three access barriers: decoding fatigue on grade-level text, written-output limitation under time pressure, and multi-step oral direction processing in math. The accommodation menu drafted above offers two to three candidates per barrier, organized by UDL category (representation, action & expression; no engagement-specific barriers this cycle).

Three recommendations carry construct-validity trade-offs the team should discuss before adoption: TTS on ELA reading, STT on ELA writing, and word-processor spell-check on CAASPP. Each is allowable as an Accommodation but flags the corresponding CAASPP claim; the team's decision is whether to accept the flag for access. The skill does not recommend silently — each is routed to team discussion.

Three recommendations require cross-coordination before the plan effective date (2026-06-12): the visual schedule / step-tracker app, the STT training, and the Read&Write predictive-text + grammar configuration all require AT-specialist consultation. Recommend scheduling Mr. Iwuoha (ext. 312, from config) by 2026-06-05.

One implementation-fidelity gap is flagged: the prior-IEP extended-time accommodation is being administered inconsistently in social studies. The team should discuss whether the issue is fidelity, scheduling, or scope misunderstanding — and what structural change would close the gap. Adding new accommodations on top of a non-faithfully-implemented one is the most-named failure pattern in the special-education accommodation literature.

One student-voice item to surface explicitly: Student M has named the resource-room pull-out as stigmatizing because of when it happens, not because of what happens. The team should discuss a schedule shift before adopting any new pull-out time.

The family has named Spanish translation of the parent-facing IEP summary; route through district translation service (Lenguaje Verde) via FACE office (from config) — both standard.

No anti-bias flags this cycle. No behavioral-control recommendations are on the menu (intentional; access barriers are academic). No accommodations are being dropped silently — each retained accommodation has a fade-plan question.

The team-discussion-question list (10 items) is the primary deliverable. The recommendation menu is the discussion material.

---

AI-use disclosure block

This recommendation menu was drafted with AI assistance for the case manager's IEP/504-meeting preparation. The case manager and the IEP/504 team are the human authors of any accommodations adopted. The team reviewed each recommendation against the student's evaluation data, present-levels summary, and family/student input before adoption. Edit history is preserved in the case manager's working file. This block is retained in the working file, not pasted into the IEP plan itself.