Skip to Content

Wiki source code of Roles & Publication Modes

Version 1.2 by Robert Schaub on 2025/12/23 17:40
Show last authors
1 = Roles & Publication Modes =
2
3 **Version:** 0.9.70
4 **Last Updated:** December 21, 2025
5 **Status:** CORRECTED - Automation Philosophy Consistent
6
7 This page defines user roles and publication modes in FactHarbor.
8
9
10 == 1. User Roles ==
11
12 === 1.1 Contributors ===
13
14 **Who**: Community members who suggest system improvements
15
16 **Responsibilities:**
17 * Participate in sampling audits (analyze patterns, not individual outputs)
18 * Suggest algorithm/prompt improvements based on findings
19 * Document systematic issues observed
20 * Contribute to system improvement discussions
21
22 **Can:**
23 * Edit published content (changes apply immediately, Wikipedia-style)
24 * Flag quality issues (for sampling audit)
25 * Earn reputation through contributions
26 * Participate in RFC (Request for Comments) processes
27
28 **Cannot:**
29 * Approve content before publication
30 * Override quality gates
31 * Act as gatekeepers
32 * Manually fix individual AI outputs
33
34 **Note**: Contributors improve THE SYSTEM, not individual outputs.
35
36
37 === 1.2 Trusted Contributors ===
38
39 **Who**: Contributors with proven track record and domain expertise
40
41 **Responsibilities:**
42 * Same as Contributors, plus:
43 * Review complex algorithm changes
44 * Provide domain expertise on contested claims
45 * Mentor new contributors
46
47 **Can:**
48 * Everything Contributors can do, plus:
49 * Participate in higher-level system design decisions
50 * Review RFC proposals
51 * Access more detailed system metrics
52
53 **Cannot:**
54 * Approve content before publication
55 * Override quality gates
56 * Act as gatekeepers
57
58 **Important**: "Trusted" refers to judgment quality, NOT approval authority.
59
60
61 === 1.3 Moderators ===
62
63 **Who**: Team members focused on community health and abuse prevention
64
65 **Responsibilities:**
66 * Handle abuse, spam, and harassment
67 * Enforce community guidelines
68 * Respond to user reports
69 * Manage bans and appeals
70
71 **Can:**
72 * Hide abusive content
73 * Ban users for policy violations
74 * Review appeals
75 * Escalate serious issues to Governing Team
76
77 **Cannot:**
78 * Approve content for publication
79 * Review content quality before publication
80 * Override quality gates for content
81 * Act as editorial gatekeepers
82
83 **Critical Distinction:**
84 ```
85 Moderators handle: ABUSE (spam, harassment, violations)
86 Moderators DO NOT handle: CONTENT QUALITY (that's automated)
87 ```
88
89
90 === 1.4 Domain Trusted Contributors (Optional, Task-Specific) ===
91
92 **Who**: Subject matter specialists invited for specific high-stakes disputes
93
94 **Not a permanent role**: Contacted externally when needed for contested claims in their domain
95
96 **When used:**
97 * Medical claims with life/safety implications
98 * Legal interpretations with significant impact
99 * Scientific claims with high controversy
100 * Technical claims requiring specialized knowledge
101
102 **Process:**
103 * Moderator identifies need for expert input
104 * Contact expert externally (don't require them to be users)
105 * Trusted Contributor provides written opinion with sources
106 * Opinion added to claim record
107 * Trusted Contributor acknowledged in claim
108
109 **Important**: This is CONSULTATION, not APPROVAL. Their opinion is added to the evidence, not used as a gate.
110
111 **User Needs served**: UN-16 (Expert validation status)
112
113
114 == 2. Publication Modes ==
115
116 **Fulfills**: UN-1 (Trust indicators), UN-16 (Review status transparency)
117
118 FactHarbor uses **TWO publication modes** (not three). Focus is on transparency and confidence scoring, not gatekeeping.
119
120 === 2.1 Mode 1: Draft-Only ===
121
122 **Status**: Not visible to public
123
124 **When Used:**
125 * Quality gates failed
126 * Confidence below threshold
127 * Structural integrity issues
128 * Insufficient evidence
129
130 **What Happens:**
131 * Content remains private
132 * System logs failure reasons
133 * Prompts/algorithms improved based on patterns
134 * Content may be re-processed after improvements
135
136 **NOT "pending human approval"** - it's blocked because it doesn't meet automated quality standards.
137
138
139 === 2.2 Mode 2: AI-Generated (Public) ===
140
141 **Status**: Published and visible to all users
142
143 **When Used:**
144 * Quality gates passed
145 * Confidence ≥ threshold
146 * Meets structural requirements
147 * Sufficient evidence found
148
149 **Includes:**
150 * Confidence score displayed (0-100%)
151 * Risk tier badge (A/B/C)
152 * Quality indicators
153 * Clear "AI-Generated" labeling
154 * Sampling audit status
155
156 **Labels by Risk Tier:**
157 * **Tier A (High Risk):** "⚠️ AI-Generated - High Impact Topic - Seek Professional Advice"
158 * **Tier B (Medium Risk):** "🤖 AI-Generated - May Contain Errors"
159 * **Tier C (Low Risk):** "🤖 AI-Generated"
160
161 **User Contributions:**
162 * User edits apply immediately (Wikipedia model)
163 * All changes logged and versioned
164 * May be selected for sampling audit
165 * Reputation earned for quality contributions
166
167
168 === REMOVED: Mode 3 ===
169
170 **V0.9.50 Decision:** No centralized approval workflow.
171
172 **Rationale:**
173 * Defeats automation purpose
174 * Creates bottleneck
175 * Inconsistent quality
176 * Not scalable
177
178 **What Replaced It:**
179 * Better quality gates
180 * Sampling audits for system improvement
181 * Transparent confidence scoring
182 * Risk-based warnings
183
184
185 == 3. Content States ==
186
187 === 3.1 Published ===
188
189 **Status**: Visible to all users
190
191 **Includes:**
192 * AI-generated analyses (default state after passing gates)
193 * User-contributed content
194 * Edited/improved content
195
196 **Quality Indicators** (displayed with content):
197 * **Confidence Score**: 0-100% (AI's confidence in analysis)
198 * **Source Quality Score**: 0-100% (based on source track record)
199 * **Controversy Flag**: If high dispute/edit activity
200 * **Completeness Score**: % of expected fields filled
201 * **Last Updated**: Date of most recent change
202 * **Edit Count**: Number of revisions
203 * **Review Status**: AI-generated / Enhanced by contributors
204
205 **Automatic Warnings:**
206 * Confidence < 60%: "Low confidence - use caution"
207 * Source quality < 40%: "Sources may be unreliable"
208 * High controversy: "Disputed - multiple interpretations exist"
209 * Medical/Legal/Safety domain: "Seek professional advice"
210
211 **User Needs served**: UN-1 (Trust score), UN-9 (Methodology transparency), ~~UN-15 (Evolution timeline - Deferred)~~, UN-16 (Review status)
212
213
214 === 3.2 Hidden ===
215
216 **Status**: Not visible to regular users (only to moderators)
217
218 **Reasons:**
219 * Spam or advertising
220 * Personal attacks or harassment
221 * Illegal content
222 * Privacy violations
223 * Deliberate misinformation (verified)
224 * Abuse or harmful content
225
226 **Process:**
227 * Automated detection flags for moderator review
228 * Moderator confirms and hides
229 * Original author notified with reason
230 * Can appeal to board if disputes moderator decision
231
232 **Note**: Content is hidden, not deleted (for audit trail)
233
234
235 == 4. Role Evolution Path ==
236
237 **Contributor Journey:**
238
239 1. **Visitor** – Explores platform, reads documentation, may raise questions
240 2. **New Contributor** – Submits first improvements (typo fixes, small clarifications, new issues)
241 3. **Contributor** – Contributes regularly and follows project conventions
242 4. **Trusted Contributor** – Has a track record of high-quality work and reliable judgment
243 5. **Auditor** – Participates in sampling audits (pattern analysis)
244 6. **Moderator** – Focuses on behavior, tone, and conflict moderation (not content quality)
245 7. **Domain Expert** (optional) – Offers domain expertise without changing governance authority
246
247 **Key Principle**: Low barrier to entry, transparent criteria for advancement, clear separation between content quality (automated) and behavior moderation (human).
248
249
250 == 5. Principles ==
251
252 * **Low barrier to entry** for new contributors
253 * **Transparent criteria** for gaining and losing responsibilities
254 * **Clear separation** between content quality (automated) and behavioral moderation (human)
255 * **Documented processes** for escalation and appeal
256 * **No gatekeeping** for content publication
257 * **Immediate application** of contributions (Wikipedia model)
258
259
260 == 6. Related Pages ==
261
262 * [[AKEL>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome]] - AI system
263 * [[Workflows>>FactHarbor.Specification.Workflows.WebHome]] - Process workflows
264 * [[Architecture>>FactHarbor.Specification.Architecture.WebHome]] - System architecture
265 * [[Decision Processes>>FactHarbor.Organisation.Decision-Processes.WebHome]] - Governance
266 * [[Requirements>>FactHarbor.Specification.Requirements.WebHome]] - All requirements
267
268
269 **V0.9.70 CHANGES:**
270
271 **REMOVED:**
272 - All references to "Mode 3" publication
273 - "Contributors validate quality gates"
274 - "Trusted Contributors validate outputs"
275 - "Moderators finalize publication"
276
277 **CLARIFIED:**
278 - Contributors suggest system improvements (not approve outputs)
279 - Moderators handle abuse only (not content quality)
280 - Trusted Contributors provide consultation (not validation)
281 - Domain experts provide opinions (not approval)
282 - Only 2 publication modes (AI-Generated / Draft-Only)
283
284 **ADDED:**
285 - Clear role boundaries
286 - Explicit "cannot" lists for each role
287 - Publication mode details
288 - Content state specifications