ceo-orchestration 1.0.0

package/.claude/skills/domains/academic-humanities/examples/PLAN-EXAMPLE-ACH.md

@@ -0,0 +1,126 @@
+---
+plan_id: PLAN-EXAMPLE-ACH
+title: "Launch oral history digital archive with IRB governance and community agreement"
+status: draft
+owner: ceo
+level: L3
+squad: academic-humanities
+profile: core,academic-humanities
+created_at: 2026-05-10
+---
+
+# Example PLAN — Oral History Digital Archive Launch
+
+> **This is an illustrative example**, not a real plan. It shows
+> how the academic-humanities squad coordinates on a digital archive
+> that touches all three VETO scopes: human-subjects data and community
+> consent (Dr. Nkosi), source attribution and editorial accuracy
+> (Prof. Marchand), and repository metadata and licensing (Gabriel Oduya).
+>
+> Exemplar pattern derived from:
+> `.claude/skills/domains/edtech/examples/PLAN-EXAMPLE.md`
+
+## 1. Problem
+
+A university humanities department is launching a publicly accessible
+digital archive of 400 oral history recordings collected between 1975-1990
+from indigenous communities in a specific region. The recordings have been
+held in physical format in the university's special collections. Making
+them accessible online raises IRB governance questions (most recordings
+predate modern consent standards), community agreement requirements, and
+archival metadata standards for long-term preservation.
+
+Sources:
+- Physical archive: 400 reel-to-reel recordings, 200 photographs, 50 field
+  journals — all held by university special collections
+- Community stakeholders: 3 indigenous community representatives with
+  interest in the materials and claim to community cultural property
+- Original researcher's notes: available but non-digitised
+
+## 2. Scope
+
+**In:**
+- IRB review for use and publication of pre-1990 oral history recordings
+  as research materials
+- Community consultation process and community research agreement
+- Digitisation and metadata creation (Dublin Core) for all 650 materials
+- Persistent identifier (ARK) assignment for all records
+- Public-facing repository with community-controlled access tiers
+
+**Out:**
+- Transcription of recordings (Phase 2 — separate project)
+- Publication of derived scholarly articles (separate manuscript workflows)
+- Digitisation of physical field journals (cost and scope constraint)
+
+## 3. Squad assignments
+
+| Phase | Owner | Deliverable |
+|---|---|---|
+| P1 — IRB and community | Dr. Amara Nkosi | IRB protocol for pre-1990 materials; community consultation plan (ACH-001) |
+| P2 — Community agreement | Dr. Nkosi + Dr. Villanueva | Community research agreement; access tier definition with communities |
+| P3 — Licensing audit | Gabriel Oduya | Copyright status of all 650 items; permissions from community (ACH-009) |
+| P4 — Digitisation + format | Takeshi Inoue | Open format digitisation (TIFF/FLAC); conversion documentation (ACH-012) |
+| P5 — Metadata creation | Gabriel Oduya + Takeshi | Dublin Core metadata for all 650 records; schema validation (ACH-011) |
+| P6 — Persistent identifiers | Takeshi Inoue | ARK assigned to every record and collection (ACH-010) |
+| P7 — Editorial review | Prof. Helene Marchand | Attribution standards for recordings; contributor credit policy (ACH-007) |
+| P8 — Repository launch | CEO + all VETO holders | Dr. Nkosi + Gabriel + Prof. Marchand sign-off |
+
+## 4. Risk axes and VETO holders
+
+- **Dr. Amara Nkosi (IRB/Ethics Coordinator):** Pre-1990 recordings were
+  collected under consent standards that may not satisfy modern IRB requirements
+  for public digital archive use → BLOCK if IRB protocol for archival use
+  has not been approved (ACH-001); BLOCK if community agreement has not been
+  executed before any recordings are made publicly accessible (ACH-002).
+- **Prof. Helene Marchand (Editor):** Original researchers and community
+  speakers must be credited per agreed attribution policy → BLOCK if contributor
+  credit and attribution policy has not been agreed with community representatives
+  before launch (ACH-007).
+- **Gabriel Oduya (Reference Librarian):** Each recording's licensing status
+  must be determined — university ownership, community cultural property, or
+  joint — before public access is granted → BLOCK if any record with
+  unresolved licensing is publicly accessible (ACH-009); BLOCK if any record
+  lacks metadata schema conformance or persistent identifier (ACH-010, ACH-011).
+
+## 5. Task chains invoked
+
+- `academic-humanities-new-research-project` — for IRB protocol and
+  community agreement phases
+- `academic-humanities-digital-archive-population` — for metadata, format,
+  identifier, and repository launch phases
+- `academic-humanities-manuscript-submission-review` — skipped for this plan
+  (applies to derived publications, not the archive itself)
+
+## 6. Acceptance
+
+- IRB protocol approved in writing for use and public access of pre-1990
+  materials (ACH-001)
+- Community research agreement executed with all 3 community representatives
+  before any recordings go public (ACH-002)
+- Copyright/licensing status confirmed for all 650 items; community cultural
+  property items restricted per community agreement (ACH-009)
+- All materials in open formats (TIFF/FLAC) with documented conversion
+  process (ACH-012)
+- Dublin Core metadata schema validation passes with 0 non-conforming records
+  (ACH-011)
+- ARK persistent identifiers assigned to all records and collections (ACH-010)
+- Attribution and contributor credit policy agreed with community before launch (ACH-007)
+- Re-identification risk assessment completed for any materials containing
+  identifiable personal information (ACH-004)
+
+## 7. Metrics
+
+- Archive accessibility: 100% of records findable via Dublin Core subject search
+- Community engagement: quarterly review meeting with community representatives
+  for the first 2 years post-launch
+- **IRB compliance audit** (annual — independent review of community agreement
+  adherence and access tier enforcement)
+
+## 8. References
+
+- `.claude/skills/domains/academic-humanities/skills/historian/SKILL.md`
+- `.claude/skills/domains/academic-humanities/skills/anthropologist/SKILL.md`
+- `.claude/skills/domains/academic-humanities/task-chains.yaml` — `academic-humanities-new-research-project`
+- `.claude/skills/domains/academic-humanities/task-chains.yaml` — `academic-humanities-digital-archive-population`
+- UNDRIP Article 31 — indigenous peoples' right to maintain, control, protect
+  and develop their cultural heritage and traditional knowledge

package/.claude/skills/domains/academic-humanities/pitfalls.yaml

@@ -0,0 +1,68 @@
+# Pitfalls Catalog — Academic-Humanities Squad
+# Loaded alongside the universal catalog when the academic-humanities profile
+# is installed. Format mirrors .claude/pitfalls-catalog.yaml:
+# id, rule, whenToUse, agents.
+
+pitfalls:
+  # === IRB AND HUMAN-SUBJECTS ETHICS ===
+  - id: ACH-001
+    rule: "NEVER collect, process, or retain human-subjects data under an expired or unapproved IRB protocol. A lapsed IRB approval is not a minor procedural issue — continuing research after lapse requires retroactive review and may require the research to be repeated. 'We're almost done' is not a justification."
+    whenToUse: "any research project that involves human participants, their data, or oral history recordings; any project that may require IRB review"
+    agents: [Dr. Amara Nkosi]
+
+  - id: ACH-002
+    rule: "Consent forms MUST NOT be changed without IRB review and approval of the amended form. A researcher who updates consent language unilaterally — even to 'clarify' or 'simplify' — has invalidated the consent for participants enrolled after the change."
+    whenToUse: "any modification to participant consent forms, information sheets, or recruitment materials"
+    agents: [Dr. Amara Nkosi]
+
+  - id: ACH-003
+    rule: "Research involving minors requires BOTH age-appropriate assent from the minor AND parental/guardian consent via an IRB-approved method. A school's blanket consent to conduct research on students does NOT substitute for parental consent under most IRB standards. Verify the specific IRB policy."
+    whenToUse: "any research involving participants under 18; any educational research that might contact students"
+    agents: [Dr. Amara Nkosi]
+
+  - id: ACH-004
+    rule: "'Anonymised' datasets MUST pass a re-identification risk assessment before release or sharing. Removing a name column is not anonymisation. Combination of zip code + birth year + gender re-identifies 87% of Americans (Sweeney 2000). Assess k-anonymity with at least the quasi-identifiers present in the data."
+    whenToUse: "any release, sharing, or publication of research data originally collected with personal identifiers"
+    agents: [Dr. Amara Nkosi, Takeshi Inoue]
+
+  # === CITATION ACCURACY AND EDITORIAL STANDARDS ===
+  - id: ACH-005
+    rule: "NEVER include a direct quote in a manuscript without consulting the original source document. A quote passed down through secondary literature may be misattributed, truncated, or contextually distorted. 'Commonly attributed to X' is not a citation — it is an admission that the source was not verified."
+    whenToUse: "any manuscript preparation that includes direct quotations; any AI-assisted writing that proposes citations"
+    agents: [Prof. Helene Marchand, Dr. Esperanza Villanueva]
+
+  - id: ACH-006
+    rule: "AI-generated citations MUST be independently verified against the actual source before inclusion. Large language models hallucinate plausible but non-existent citations. 'The AI generated it' is not a defence against a retraction for fabricated sources."
+    whenToUse: "any use of AI writing or research assistants for literature review or citation generation"
+    agents: [Prof. Helene Marchand, Gabriel Oduya]
+
+  - id: ACH-007
+    rule: "Authorship MUST be assigned based on contribution criteria (e.g. ICMJE: conception, design, execution, interpretation AND draft review AND final approval). Seniority, funding source, and lab head status do not constitute authorship. Gift and ghost authorship are both violations of academic integrity."
+    whenToUse: "any manuscript preparation; any research output where authorship is being assigned"
+    agents: [Prof. Helene Marchand]
+
+  - id: ACH-008
+    rule: "Citation spot-checks (minimum 20% of citations verified against primary source) are REQUIRED before submission. A manuscript submitted without a citation audit is a manuscript with unverified claims. The editor's review is not a substitute for the author's citation verification."
+    whenToUse: "any manuscript at pre-submission stage; any revised manuscript before resubmission"
+    agents: [Prof. Helene Marchand]
+
+  # === ARCHIVE ACCESS AND REPOSITORY STANDARDS ===
+  - id: ACH-009
+    rule: "Archival images, documents, and digitised materials MUST have confirmed licensing from the holding institution before publication. 'Old' does not mean public domain — copyright term varies by jurisdiction, author death date, and first-publication status. For materials held by private archives or families, always obtain written permission."
+    whenToUse: "any publication or digital output that includes reproductions of archival materials"
+    agents: [Gabriel Oduya]
+
+  - id: ACH-010
+    rule: "Research datasets cited in publications MUST include a version number and persistent identifier (DOI, ARK, or PURL). A URL that resolves to a frequently-updated dataset is not a reproducible citation — the dataset the paper relied on may no longer exist at that URL."
+    whenToUse: "any data citation in a manuscript; any release of a shared research dataset"
+    agents: [Gabriel Oduya, Takeshi Inoue]
+
+  - id: ACH-011
+    rule: "Repository metadata MUST conform to the declared standard (Dublin Core, EAD, MARC21) for every record — not just a majority. Incomplete or non-conforming metadata creates unfindable records. 'We'll fix the metadata later' means 40% of the collection is effectively invisible to researchers."
+    whenToUse: "any repository population event; any metadata migration; any new record type being added to an existing repository"
+    agents: [Gabriel Oduya, Takeshi Inoue]
+
+  - id: ACH-012
+    rule: "Research data MUST be stored in open, non-proprietary formats and archived with version control before any analysis begins. Data stored only in a local folder, a proprietary tool, or a single researcher's machine is data that will be lost when the project ends. Minimum: open format + version-controlled repository + institutional backup."
+    whenToUse: "any new research project data infrastructure design; any computational analysis pipeline; any project handoff"
+    agents: [Takeshi Inoue, Dr. Esperanza Villanueva]

package/.claude/skills/domains/academic-humanities/skills/anthropologist/SKILL.md

@@ -0,0 +1,394 @@
+---
+name: anthropologist
+description: |
+  Anthropological lens for product, market, and organizational research.
+  Covers ethnographic fieldwork design, participant observation, emic/etic
+  perspective management, thick description (Geertz), cultural relativism,
+  and the application of kinship, ritual, and exchange frames to user research
+  and team dynamics. Applies reflexivity discipline throughout: the researcher's
+  positionality is acknowledged and surfaced, not hidden. Enforces IRB-equivalent
+  informed-consent standards and LGPD/GDPR compliance for personal-data
+  ethnography. Use when: designing a user research study that requires deep
+  contextual understanding; interpreting behavioral data that resists survey
+  explanation; analysing organizational dysfunction through ritual or exchange
+  frames; evaluating whether product design imposes etic categories on users
+  whose emic models differ; or reviewing research deliverables for extractive
+  or romanticizing framing.
+owner: Anthropologist (domain persona)
+tier: domain:academic-humanities
+scope_tags: [anthropology, ethnography, participant-observation, thick-description, cultural-relativism, user-research]
+inspired_by:
+  - source: msitarzewski/agency-agents/academic/academic-anthropologist.md@783f6a72bfd7f3135700ac273c619d92821b419a
+    license: MIT
+    relationship: structural_inspiration
+    authored_by: ceo-orchestration framework
+    authored_at: 2026-05-07
+# --- smart-loading fields (PLAN-083 Wave 0b sub-agent 0.7c) ---
+domain: academic-humanities
+priority: 8
+risk_class: low
+stack: []
+context_budget_tokens: 500
+inactive_but_retained: true
+repo_profile_binding:
+  frontend: {active: false, priority: 10}
+  engine: {active: false, priority: 10}
+  fintech: {active: false, priority: 10}
+  trading-readonly: {active: false, priority: 10}
+  generic: {active: false, priority: 10}
+activation_triggers: []
+# --- K1 paths: native file-touch activation (PLAN-135 W3 unit k1a) ---
+paths:
+  - "**/research/**"
+  - "**/user-research/**"
+  - "**/fieldwork/**"
+---
+
+# Anthropologist
+
+## Cardinal Rule
+
+No ethnographic deliverable may assert a cultural explanation without both
+a documented field-basis (observation log, interview record, or published
+secondary source) AND an explicit reflexivity statement acknowledging how
+the researcher's positionality may shape the interpretation. Reports that
+present cultural claims as objective facts without these two elements are
+rejected at the two-pass review gate (ADR-058).
+
+## Fail-Fast Rule
+
+Stop and return a structured failure when any of the following is true:
+
+- Informed consent has not been obtained from research participants, or
+  the consent process did not adequately communicate the study purpose,
+  data use, and right to withdraw.
+- Research involves a vulnerable population (minors, incarcerated persons,
+  persons in dependent relationships with the researcher) without an
+  explicit protocol approved by the responsible governance body.
+- Personal-data collection would constitute processing under LGPD (Lei
+  13.709/2018) or GDPR without a documented lawful basis and data-minimisation
+  justification.
+- The researcher has a material conflict of interest with the community
+  being studied that has not been disclosed to participants.
+- The study design requires deception and no deception-specific consent
+  protocol has been prepared.
+
+Never approximate fieldwork. Thin observation periods produce findings
+with undocumented confidence limits, not approximated certainty.
+
+## When to Apply
+
+Apply this skill when:
+
+- Designing a user research study for a product or service in a context
+  where cultural framing materially affects behaviour (adoption barriers,
+  taboo objects, gift vs. transaction norms, authority and permission structures).
+- Interpreting qualitative findings that resists reduction to survey metrics —
+  when the "why" behind observed behaviour requires contextual thickness.
+- Analysing organizational dynamics through structural lenses: onboarding as
+  rite of passage, reporting structures as kinship, reciprocity norms in team
+  economy.
+- Evaluating whether a product's conceptual model (information architecture,
+  naming, workflows) imposes etic categories on users whose emic mental models
+  differ.
+- Reviewing research deliverables — internal reports, personas, journey maps —
+  for extractive framing, romantic othering, or under-acknowledged positionality.
+
+Do not apply this skill as a substitute for quantitative methods when the
+research question requires statistical power. Ethnographic and survey methods
+are complementary; route to `domains/academic-humanities/skills/psychologist`
+for psychometric instrument design.
+
+## Ethnographic Method
+
+Participant observation is the primary instrument of ethnographic inquiry.
+Surveys and focus groups are supplementary and are treated as data on
+self-reported behaviour, not on behaviour itself.
+
+### Field-Time Investment
+
+Thin ethnography — a single session, a short survey with open-text fields,
+or a few hours of observation — produces hypotheses, not findings. Document
+the observation period explicitly in every deliverable. The minimum threshold
+for treating observation as evidential varies by context:
+
+- Familiar cultural context (researcher shares background with participants):
+  multiple sessions spanning at least two distinct activity cycles relevant
+  to the research question.
+- Unfamiliar cultural context: extended immersion; the researcher should
+  reach a point of diminishing interpretive surprise before claiming saturation.
+
+When field time is constrained by project conditions, the deliverable must
+state the limitation and qualify findings accordingly. Overstating confidence
+from thin observation is the primary validity failure mode in applied ethnography.
+
+### Emic vs. Etic Perspective
+
+Emic perspective: how participants categorize, explain, and give meaning to
+their own practices. Always establish the emic frame first.
+
+Etic perspective: analytical categories the researcher applies from outside
+the community — theoretical frameworks, comparative taxonomies, design heuristics.
+Etic frames are legitimate analytical tools, but they must be applied consciously
+and never presented to participants as if they were the community's own categories.
+
+The error pattern to prevent: importing product-team vocabulary (e.g., "user
+journey", "pain point", "delight") into fieldwork instruments before the emic
+vocabulary has been established. This contaminates data at the point of collection.
+
+## Thick Description
+
+Thick description (Geertz, "The Interpretation of Cultures", 1973) is the
+practice of interpreting cultural acts by restoring the layered context —
+institutional, historical, relational — that makes the act meaningful to
+participants.
+
+### Geertz Framework Applied
+
+A thin description of an observed act records the surface event: "the team
+lead rejected the junior engineer's code". A thick description records the
+act within its web of significance: the reporting structure, the history of
+prior reviews, the team's shared norm about public correction, the junior
+engineer's probationary status, and the institutional pressure from the
+upcoming release — each layer transforming the interpretive meaning of the
+same surface event.
+
+Procedure for producing thick description:
+
+1. Record the surface event with precision (who, what, when, where).
+2. Elicit participant accounts of the event's meaning — do not infer meaning
+   from observation alone.
+3. Map the institutional and relational context in which the event is embedded.
+4. Identify any symbolic registers the act activates (status, reciprocity,
+   authority, membership, exclusion).
+5. State interpretive claims as interpretations, not facts. Use "this act
+   appears to function as..." rather than "this act means...".
+
+Never strip context in the name of concision. A finding that cannot survive
+the context stripping required to fit a slide deck should be presented with
+the context, not condensed into a decontextualized claim.
+
+## Cultural Relativism Discipline
+
+Cultural relativism is a methodological commitment during the research phase:
+suspend evaluative judgment to allow accurate description. It is not a
+normative claim that all cultural practices are equally defensible.
+
+The distinction matters operationally:
+
+- During fieldwork and analysis: treat observed practices as solutions to
+  problems, not as deviations from a norm. Ask "what problem does this practice
+  solve for these people?" before evaluating whether the practice is optimal.
+- During reporting and recommendation: evaluative judgment is appropriate and
+  necessary, but it must follow description, not precede it. Judgments formed
+  before adequate description reflect researcher bias, not analytical finding.
+
+Cultural relativism does not require neutrality on practices that cause harm.
+The discipline is to describe accurately first; evaluation informed by that
+description comes second.
+
+## Kinship, Ritual, Exchange Frames
+
+Structural frames from social anthropology apply directly to organizational
+and product contexts. Apply these frames when standard organizational analysis
+produces thin explanation.
+
+### Kinship Frame — Reporting Structures
+
+Kinship analysis asks: who is obligated to whom, and on what basis? Applied
+to organizations:
+
+- Formal reporting structures map approximately to descent: who inherits
+  authority, resources, and obligations from whom.
+- Informal networks of mutual obligation (mentorship, sponsorship, coalition)
+  map to fictive kinship — relationships that carry kinship-like obligations
+  without formal organizational sanction.
+- Inheritance conflicts (succession tension, ownership disputes, rival factions)
+  are structurally isomorphic to kinship disputes in classic ethnographic
+  literature.
+
+Diagnostic value: when formal authority and informal obligation diverge,
+map both. The divergence is typically where organizational dysfunction lives.
+
+### Ritual Frame — Onboarding and Ceremony
+
+Ritual analysis (van Gennep, "The Rites of Passage", 1909; Turner, "The
+Ritual Process", 1969) identifies three phases: separation, liminality,
+incorporation.
+
+Applied to organizational onboarding:
+
+- Separation: the new member is marked as distinct from existing members
+  (badge colour, restricted access, formal introduction ceremonies).
+- Liminality: the threshold period where the new member is neither fully
+  outside nor fully inside — probationary status, supervised access,
+  assigned buddy or mentor. This phase carries anxiety and is structurally
+  generative: identity is malleable.
+- Incorporation: the act or event that signals full membership — first
+  independent decision, public acknowledgment by senior members, removal
+  of restrictions.
+
+When onboarding dysfunction is reported, map it against these three phases.
+Failure is typically located in one phase: inadequate separation (new members
+begin without clear role definition), prolonged liminality (no clear incorporation
+signal), or absent incorporation (new members remain in indefinite threshold status).
+
+### Exchange Frame — Reciprocity in Team Economy
+
+Exchange analysis (Mauss, "The Gift", 1925; Polanyi, "The Great Transformation",
+1944) identifies three modes: reciprocity, redistribution, and market exchange.
+
+Applied to team dynamics:
+
+- Reciprocity: informal mutual aid and knowledge sharing operate on gift-economy
+  logic. Contributions are not immediately repaid but create ongoing obligation.
+  When reciprocity norms break down (one-way knowledge extraction, credit
+  appropriation, free-riding), team cohesion degrades in ways that survey instruments
+  measure as "culture problems" without locating the structural cause.
+- Redistribution: centralized allocation of resources, access, and recognition
+  through a manager or coordinator. Perceived fairness of redistribution is a
+  primary driver of legitimacy.
+- Market exchange: explicit, time-bounded, transactional. When team members
+  begin treating internal relationships as market exchanges — demanding explicit
+  credit before contributing, time-boxing help to billable-equivalent units —
+  the exchange-mode shift signals a breakdown in reciprocity norms.
+
+## Reflexivity Practice
+
+Reflexivity is the practice of acknowledging how the researcher's position,
+background, prior knowledge, and institutional affiliation shape both the
+data collection and the interpretation.
+
+Reflexivity is not an optional caveat. In applied ethnography, un-surfaced
+positionality is a validity threat, not a stylistic concern.
+
+Procedure for reflexivity documentation:
+
+1. Before fieldwork: write a positionality statement documenting the
+   researcher's relevant affiliations, assumptions, and prior expectations
+   about the community or context. This statement is filed with the research
+   plan, not suppressed.
+2. During analysis: maintain an analytic memo tracking interpretive choices
+   and the reasoning behind them. When alternative interpretations were
+   considered and rejected, document why.
+3. In the deliverable: include a reflexivity section (brief — two to four
+   paragraphs) that summarises the researcher's position and notes any
+   specific ways it may have shaped the findings.
+
+Required fields for the positionality statement filed with every research plan:
+
+```
+Positionality Statement — [Study Title]
+────────────────────────────────────────────────────────
+researcher_affiliation:      [Institution or team; relationship to sponsor]
+prior_knowledge_of_context:  [Relevant experience with this community or domain]
+expected_findings_before_fieldwork: [Stated hypotheses or assumptions held
+                                      before observation began]
+potential_bias_vectors:      [Specific ways affiliation or prior knowledge may
+                               shape data collection or interpretation]
+mitigation_approach:         [How each bias vector will be managed in fieldwork
+                               design and in analysis — or "None identified"]
+────────────────────────────────────────────────────────
+reflexivity_section_included_in_deliverable: yes / no (if no: documented rationale)
+```
+
+Omitting the positionality statement is a blocker finding at the two-pass
+review gate (ADR-058).
+
+The reflexivity section is not a disclaimer. It is evidence that the
+researcher has engaged in the intellectual discipline required for valid
+interpretation.
+
+## IRB and Ethical Compliance
+
+### Informed Consent
+
+Informed consent requires that participants understand: the study's purpose,
+what data will be collected, how data will be stored and used, who will
+have access to findings, and that participation is voluntary and
+withdrawal is possible at any time without penalty.
+
+Consent forms must use language accessible to the participant population.
+Technical or legal language that obscures rather than communicates does not
+constitute valid informed consent.
+
+For longitudinal studies: re-consent at each phase if the scope or use of
+data changes from what was originally described.
+
+### Vulnerable Populations
+
+Research involving minors, persons in custodial or institutional settings,
+individuals in economically dependent relationships with the researcher's
+institution, or communities with historical experience of extractive research
+requires:
+
+- A protocol reviewed by an ethics or governance body before fieldwork begins.
+- Additional protections specified in the protocol (e.g., parental consent
+  for minors, independent advocate for participants in dependent relationships).
+- Particular attention to whether voluntary participation is genuinely achievable
+  given power differentials.
+
+### Data Sovereignty — Indigenous Research
+
+Research with indigenous communities requires engagement with the principle
+of data sovereignty: the community's right to govern research conducted on
+or about them, including rights over data storage, access, and publication.
+
+Specific obligations vary by jurisdiction and community protocol. The
+researcher's obligation is to determine what protocols apply before fieldwork
+begins, not after data collection.
+
+### LGPD and GDPR
+
+Ethnographic data that includes personal information about identifiable
+individuals constitutes personal-data processing under LGPD (Lei 13.709/2018,
+Brazil) and GDPR (EU 2016/679).
+
+Minimum requirements:
+
+- Document the lawful basis for processing before data collection begins.
+- Apply data minimisation: collect only what the research question requires.
+- Anonymise or pseudonymise data as early in the workflow as possible.
+- Define and document retention periods; do not retain identifiable data
+  beyond research necessity.
+- Participants' rights (access, correction, deletion, portability) apply
+  to research data unless a specific research exemption is documented.
+
+## Anti-patterns
+
+| Anti-pattern | Description | Correct Approach |
+|---|---|---|
+| Extractive ethnography | Treating the community as a data source without reciprocity — publishing findings without community review, taking without giving back | Build reciprocity into the research design; offer findings to the community before external publication; ask what the community needs from the research |
+| Romantic othering | Presenting the studied community as exotic, pure, or more authentic than the researcher's own — implicitly elevating them as a counter-example to modernity | Treat the community as a complex adaptive system with its own politics, contradictions, and innovations; resist the "noble alternative" framing |
+| Generalising from N=1 | Treating a single informant or observation as representative of the community or cultural pattern | Document the basis for any generalisation; single informants provide hypotheses, not findings; triangulate across multiple sources |
+| Ignoring positionality | Presenting analysis as if researcher neutrality is achievable; suppressing the reflexivity statement | Write the positionality statement before fieldwork; include reflexivity section in all deliverables |
+| Etic imposition | Applying analytical categories derived from theory or product-team vocabulary before establishing the emic frame | Establish emic categories first through open-ended observation and participant accounts; apply etic frames explicitly and secondarily |
+| Thin description | Reducing contextually complex acts to surface-level events stripped of relational and institutional meaning | Apply the Geertz thick-description procedure; restore context layers before stating interpretive claims |
+| Citing unreviewed claims | Asserting ethnographic facts without a documented field-basis or peer-reviewed secondary source | Every cultural claim must cite its basis: observation log with date and duration, interview record with consent status, or published source with full citation |
+
+## Cross-References
+
+- `domains/academic-humanities/skills/psychologist` — for research questions
+  requiring psychometric instruments, validated scales, or statistical inference
+  from survey data; ethnographic and psychometric methods are complementary,
+  not substitutes.
+- `domains/academic-humanities/skills/narratologist` — for analysis of the
+  narrative structures through which participants explain their practices;
+  thick description and narrative analysis frequently overlap in interview data.
+- `core/code-review-checklist` — apply the two-pass review gate (ADR-058)
+  to ethnographic research deliverables before circulation; reports with
+  cultural or behavioural claims carry reputational and ethical risk equivalent
+  to high-stakes analytical artifacts.
+
+## ADR Anchors
+
+- **ADR-058 (Brainstorm gate + two-pass adversarial review):** ethnographic
+  reports, research memos, and persona documents produced under this skill
+  are primary analytical artifacts requiring the two-pass review defined in
+  ADR-058 §BORROW-2. The first pass reviews completeness and internal
+  consistency of the fieldwork basis. The second pass reviews from an
+  adversarial frame: specifically challenging whether emic/etic distinctions
+  have been maintained, whether positionality has been adequately surfaced,
+  and whether any generalisation exceeds the evidential basis. The deliverable
+  must not be circulated until both passes are complete and any blocker
+  findings are resolved.