Run bounded Ofsted Essex delay ingestion slice

[happygeneralist]

Purpose

Run the next bounded evidence-led ingestion slice from the partially ingested Essex Ofsted/CQC Area SEND inspection source.

This is a worked slice under #86. It should use real ingestion difficulty to test and refine the public guardrails in #87 without pausing ingestion for a complete operating model.

Source

Source archive:

happygeneralist/civic-design-secondary-research
sources/SRC_001_ofsted_essex_area_send_inspection_jan_2026/

Primary candidate:

Extract 002 in extracts.md

Primary source location:

Page 2, section: What is it like to be a child or young person with SEND in this area?

Limited corroborating context may be inspected on page 5 where the report discusses EHC plans issued outside the statutory timeframe and source-attributed contributors to delay.

Slice boundary

Working slice name:

Delays in EHC needs assessments, EHC plans, provision, and support

The working name does not imply that these are one pain point or one causal chain.

Included claims

Review and separate the source's claims about:

1. delays in assessment of needs;
2. poor timeliness of EHC plans;
3. some children and young people waiting too long for appropriate provision;
4. some children and young people waiting too long for support from some services;
5. family dissatisfaction associated by the report with these delays;
6. inconsistent experiences and outcomes as inspection synthesis, with the exact claim boundary preserved;
7. the page 5 EHC-plan timeframe material only where it helps clarify the plan-delay claim or the source's own explanation.

Excluded claims

Do not expand this slice to include:

• neurodevelopmental assessment waits as a whole;
• unequal access to health services by locality;
• EHC plan quality or professional input;
• communication while waiting, which has already been ingested through the first slice;
• families needing to push hard, mediation, tribunal, Ombudsman complaints or wider escalation burden;
• preparation for adulthood;
• broad leadership or governance findings;
• national prevalence, national severity or cross-locality claims;
• a general claim that one delay causes all later delays or outcomes.

Nearby objects to inspect

Before proposing an object change, inspect at least:

• PP_002_parent_carries_pathway_navigation_burden;
• UN_015_need_to_coordinate_next_steps;
• UN_007_reconcile_pathway_information;
• deprecated UN_010_consider_steps_rights_adjustments as lineage only;
• BEH_009_chases_pathway_support;
• existing insights and pain points for any genuine delay-related fit;
• current civic-need material and theory work without creating a civic need by default.

A light search has not identified an existing delay-specific pain point. That does not authorise creating one before candidate evidence review.

Questions to test

Use the slice to examine:

• whether assessment delay, plan delay, provision delay and support delay are distinct conditions;
• what the report directly supports and what is inspection synthesis;
• whether any evidence fits an existing pain point directly, partially or only contextually;
• whether a new pain point is genuinely necessary;
• whether the available evidence supports an action-oriented user need, or only points towards one for later research;
• whether the material indicates a civic condition without yet supporting a mature civic need;
• how EHC assessment and plan mechanisms should remain visible as context without defining an enduring need;
• whether evidence supports existence, manifestation, consequence, severity, prevalence, association or causation;
• which conclusions should remain deferred or needs_review.

Expected evidence types

Candidate entries may include:

• direct inspection finding;
• inspection synthesis;
• source-attributed consequence or association;
• statutory-mechanism context;
• source-asserted contributing factor;
• contextual or limiting evidence.

The table must distinguish these rather than treating all report statements as equivalent evidence.

Candidate object types

Consider, without presuming creation:

• evidence only;
• existing pain-point link or qualification;
• new narrow pain point;
• candidate user need;
• candidate insight;
• civic-condition question for later theory work;
• mechanism or service context;
• defer / inbox material.

Do not turn a statutory duty, administrative stage, plan artefact or inspection recommendation into a user need.

Public/private method boundary

Public work may contain:

• the source boundary;
• precise source locations and faithful extracts or paraphrases;
• public-safe claim classification;
• nearby-object reconciliation;
• uncertainty and deferred conclusions;
• create, link, update, challenge or defer recommendations.

Do not publish:

• protected prompt chains;
• detailed scoring or review rubrics;
• ontology-navigation logic;
• reusable private orchestration routines;
• private scratch reasoning not needed to inspect the public result.

Route a genuinely reusable public modelling problem to #87. Route a protected judgement test to happygeneralist/design-intelligence-orchestration#10 only when this slice exposes a concrete need for it.

Stop or shrink conditions

Stop or narrow the slice if:

• the candidate set grows beyond 5 to 8 useful evidence entries;
• assessment, plan, provision and support delay cannot be kept analytically distinct;
• provision and support cannot be separately located or faithfully attributed in the source;
• the work starts importing neurodevelopmental, health-equity, plan-quality or escalation material;
• a proposed user need merely restates receive an EHC plan on time or another current mechanism;
• a new pain point is being created before nearby-object reconciliation;
• the analysis begins to imply prevalence, severity or causation not supported by this local inspection source.

Fallback boundary:

Assessment-of-needs delay and EHC-plan timeliness only, with provision and support waits retained as one source-bundled deferred claim.

Acceptance criteria

• The exact source boundary and included/excluded claims are recorded.
• Assessment, plan, provision and support delays are not automatically collapsed.
• A candidate evidence table is created before knowledge-object changes.
• Each entry records precise location, faithful wording, claim type, actor and context, direct support, non-support, nearby-object fit, recommendation, confidence and uncertainty.
• The table distinguishes existence, manifestation, consequence, severity, prevalence, association and causation where relevant.
• Nearby user needs, pain points, behaviours, civic needs and insights are inspected.
• Statutory mechanisms remain context rather than canonical user-need wording.
• No object is marked reviewed or validated.
• No national or cross-locality claim is made from this source.
• Any object change is deferred to a later small PR after candidate evidence review.
• The secondary-research evidence map is updated only when selected CDI evidence objects are created.
• Concrete workflow findings are routed to Develop safer repeatable non-linear ingestion workflow #86, Define evidence-context-object and non-linear journey modelling guardrails #87, orchestration Vary public page section layout #10 or the source archive rather than spawning unnecessary issues.

Expected PR sequence

1. Small slice-plan PR linked to this issue.
2. Candidate-evidence-table PR with no knowledge-object changes.
3. Selected evidence-record PR, including the source evidence-map update.
4. Optional object-maintenance PR only for changes justified by the reviewed candidate set.
5. Optional small guidance correction only where repeated or material difficulty exposes a concrete operating need.

Validation effect

This issue authorises planning and reviewable proposals only. It does not authorise semantic object creation, maturity upgrades, merge, closure or other finalising action.

Related work

• Develop safer repeatable non-linear ingestion workflow #86 Develop safer repeatable non-linear ingestion workflow
• Define evidence-context-object and non-linear journey modelling guardrails #87 Define evidence-context-object and non-linear journey modelling guardrails
• Test ingestion of an assumptive journey analysis as a provisional view #89 Test ingestion of an assumptive journey analysis as a provisional view
• PR Review Essex evidence-to-object fit and model-led follow-up #88 Ofsted Essex evidence-to-object fit review
• happygeneralist/design-intelligence-orchestration#10
• happygeneralist/Happygeneralist#11

[happygeneralist]

Initial bounded-ingestion work is now split into reviewable stages:

• PR Plan bounded Ofsted Essex delay slice #93 — proposed slice plan and exact source boundary
• PR Review candidate evidence for Ofsted Essex delay slice #94 — candidate evidence review only

The candidate review proposes six new evidence records, reuse of existing EVID_003 rather than duplication, and deferral of the educational-psychologist recruitment claim as separate structured evidence unless later analysis needs it.

No evidence or knowledge objects have been changed. No user need, civic need, pain point or insight is proposed for creation at this stage.