architecture-decision-record-author

name	architecture-decision-record-author
description	Author one or more Architecture Decision Records (adr@1) from an approved SRS plus architectural-spike@1 output (produced by architectural-spike-author; lean profile with no spike: the ADR options table carries the evidence inline). Use when user asks to "draft an ADR", "record this architecture decision", or "document the technical choice". Captures significant technical decisions with context, options considered, decision, consequences. Aligned with Michael Nygard's ADR format and arc42 documentation conventions. Covers SDP §2(d) System architecture. Chains naturally into architecture-decision-record-audit, threat-model-author, software-design-document-author. Do NOT use for "audit this existing ADR" (use architecture-decision-record-audit instead).
license	Apache-2.0
metadata	{"version":"1.0.0","module":"skill","stage":"d","cyberos-template":"architecture-decision-record@1","cyberos-rubric-target":"adr_rubric@1.0"}
allowed_memory_scopes	{"read":["project:*","module:*","company:locked-decisions","memories:decisions","memories:projects","memories:refinements"],"write":["project:*","memories:projects"]}
allowed_mcp_tools	["kb.read","kb.search","memory.search","memory.write_memory","audit.append","chat.notify"]
escalation	{"to_persona_on_legal":"cuo-clo","to_persona_on_security":"cuo-cseco","to_persona_on_compliance":"cuo-clo","to_human_on_irreversible":true}
invocation_modes	["standalone","chained"]
expects	{"schema_ref":"./envelopes/input.json","required_fields":["source_files","output_dir"],"optional_fields":["manifest_path","batch_size","caller_persona","trace_id","chain_to"],"standalone_interview_ref":"./STANDALONE_INTERVIEW.md"}
produces	{"schema_ref":"./envelopes/output.json","output_kind":"artefact","human_summary_ref":"./HUMAN_SUMMARY.md"}
depends_on_contracts	[{"id":"adr","version":"v1","purpose":"generation_skeleton","pin_path":"cyberos/skill/contracts/architecture-decision-record/"},{"id":"nats-subjects","version":"v1","purpose":"wire_protocol_emission","pin_path":"cyberos/skill/contracts/nats-subjects/"}]
exposable_as	{"internal":true,"agent_plugin":true,"mcp_tool":true,"partner_connector":false}
audit	{"emit_to":"genie.action_log","row_kind":"artefact_write","payload_hash_field":"artefact_hash","explanation_pane":"required"}
confidence_band	{"default":0.7,"defer_below":0.5,"cite_sources":"required"}
untrusted_inputs	{"wrap_in_marker":"untrusted_content","injection_scan":"required","on_marker_hit":"surface_to_human"}
self_audit	{"invariants_ref":"./INVARIANTS.md","check_at":["on_node_boundary",{"on_audit_row_count":25},"on_completion"],"anomaly_signals":{"confidence_low_streak":{"threshold":3,"window":10},"user_correction_streak":{"threshold":2,"window":5},"denylist_near_miss_streak":{"threshold":2,"window":20},"scope_rejection_streak":{"threshold":1,"window":1},"citation_missing_streak":{"threshold":2,"window":10}},"on_breach":{"emit":"refinement_proposal","pause_pipeline":true,"resume_token_field":"refinement_run_id"}}
human_fine_tune	{"fine_tuner_role":"cpo","review_required":{"on_minor_bump":false,"on_major_bump":true,"on_safety_change":true},"signals_to_initiate":[{"acceptance_rate_below":0.6},{"hitl_pause_rate_above":0.4},{"drift_signal_count_above":3},"user_complaint_received","regulator_inquiry_received",{"self_audit_refinement_proposal_count_above":2}],"procedure_ref":null,"required_artifacts":["changelog_entry","acceptance_test_added","memory_refinement_entry"],"blackout_windows":[]}
determinism	{"reproducible":false,"fixity_notes":"Authoring is judgement; manifest state IS reproducible. Re-running on settled state is a no-op except for last_audit_at refresh."}
emitted_source_freshness_tier	30
gated_until_phase	null
untrusted_content_wrapping	required

architecture-decision-record-author — ADR generator

Standalone trigger that turns one or more documents into a versioned, audited architecture-decision-record@1 markdown. Halts at PLAN approval and HITL gates; resumable from manifest.json state. Chains naturally into architecture-decision-record-audit by default.

prompt_revision: adr_author@1.0.0

When to invoke this skill

CUO routes a request here when the user wants to:

• "Turn this into a ADR."
• "Generate v2 of the ADR from the updated source."
• "Draft a ADR for ."

If the user asks to audit an existing ADR, route to architecture-decision-record-audit instead. If both, the supervisor chains the two.

Self-test preamble — emit BEFORE any file action

Begin every invocation with a single fenced CONTRACT_ECHO block. Do NOT proceed past this block until it has been emitted.

CONTRACT_ECHO
skill_id:                        architecture-decision-record-author
skill_version:                   1.0.0
prompt_revision:                 adr_author@1.0.0
template_version:                architecture-decision-record@1   (loaded from cyberos/skill/contracts/architecture-decision-record/template.md)
output_dir:                      <from caller>
manifest_path:                   <from caller; default: <output_dir>/manifest.json>
naming_pattern:                  ADR-{NNN}-{slug}.md
batch_size:                      <from caller; default 3, soft-cap 5, hard-cap 10>
hitl_categories:                 [<list per skill — e.g. customer_quotes, success_metric_targets, scope_decomposition>]
hitl_policy:                     HALT_BATCH_ON_PAUSE
amendment_policy:                ACCUMULATE_THEN_BATCH
max_iterations_per_artefact:     10
re_entrancy:                     idempotent_on_manifest_state
untrusted_content_handling:      spotlight_xml_tagged
file_scope:                      MUST NOT write outside output_dir
inputs:
  source_files:                  [<list of paths/URLs with media_type>]
  source_hash:                   <sha256 of normalized concat, see references/MANIFEST_SCHEMA.md §3.1>
phase:                           <PLAN | WORKER | RESUME>   (computed per §3 below)

§1 Pipeline interface (envelopes)

Input envelope (envelopes/input.json):

{
  "source_files":