@forwardimpact/basecamp 2.6.1 → 2.8.0

package/template/.claude/skills/extract-entities/SKILL.md

@@ -58,6 +58,9 @@ Run this skill:
 - `knowledge/Organizations/*.md` — organization notes
 - `knowledge/Projects/*.md` — project notes
 - `knowledge/Topics/*.md` — topic notes
+- `knowledge/Roles/*.md` — role/requisition files (created or enriched)
+- `knowledge/Candidates/*/brief.md` — candidate briefs (enriched with inferred
+  metadata)
 - `~/.cache/fit/basecamp/state/graph_processed` — updated with newly processed
   files
 
@@ -385,6 +388,75 @@ Log state changes in activity with `[Field → value]` notation:
 - **2025-01-20** (email): Leadership approved pilot. [Status → active]
 ```
 
+## Step 7b: Detect Recruitment Signals
+
+When processing emails and calendar events, look for signals that relate to the
+recruitment pipeline. These signals enrich `knowledge/Roles/` and
+`knowledge/Candidates/` with metadata that cannot be derived from any single
+source.
+
+### Requisition Number Detection
+
+Scan email subjects and bodies for requisition numbers (e.g. 7-digit Workday
+IDs). When found:
+
+1. Check if a Role file exists: `ls knowledge/Roles/ | grep "{req_number}"`
+2. If **no Role file exists**, create a stub (see `track-candidates` Step 0b
+   for the template). Search the knowledge graph for context:
+   ```bash
+   rg "{req_number}" knowledge/
+   ```
+3. If a Role file **does exist**, check if the email provides new metadata
+   (hiring manager, recruiter, locations) and update the Role file.
+
+### Hiring Manager Inference from Calendar Events
+
+When a calendar event title matches interview-related patterns:
+- "Interview", "Screening", "Screen", "Decomposition", "Panel", "Technical
+  Assessment", "Candidate"
+- Combined with a person name (cross-reference `knowledge/Candidates/`)
+
+Extract the **organizer** of the event. If the organizer is NOT the user (from
+`USER.md`), they are likely the hiring manager for this role. To confirm:
+
+1. Look up the organizer in `knowledge/People/` — check if they have a role
+   indicating they manage a team or are described as a hiring manager.
+2. Look up which candidate is being interviewed — check their `brief.md` for
+   a `Req` field.
+3. If a Req is known, update the corresponding `knowledge/Roles/*.md` file's
+   `Hiring manager` field (only if currently `—`).
+4. Update the candidate's `brief.md` `Hiring manager` field if currently `—`.
+
+### Recruiter Inference from Email Threads
+
+When processing email threads that reference candidates (by name match against
+`knowledge/Candidates/`), check the To/CC fields for internal recruiters:
+
+1. Cross-reference To/CC addresses against `knowledge/People/` notes.
+2. If a CC'd person's note mentions "recruiter", "talent acquisition", or a
+   similar recruiting role, they are likely the internal recruiter for this
+   candidate's role.
+3. Update the candidate's `brief.md` recruiter field and the corresponding
+   Role file if the field is currently `—`.
+
+### Domain Lead Resolution
+
+When a hiring manager is newly identified (from calendar or email inference),
+attempt to resolve the domain lead:
+
+1. Read the hiring manager's People note for a `**Reports to:**` field.
+2. Walk up the reporting chain until reaching a VP or senior leader listed in
+   a stakeholder map or organizational hierarchy note.
+3. Update both the Role file's `Domain lead` and the candidate brief's
+   `Domain lead`.
+
+**Be conservative:** Only set hiring manager/domain lead/recruiter when the
+evidence is strong. A single calendar invite organized by someone is suggestive
+but not conclusive — confirm against People notes or multiple data points before
+setting the field.
+
+---
+
 ## Step 8: Check for Duplicates
 
 Before writing:
@@ -462,4 +534,8 @@ Before completing, verify:
 - [ ] Open items are commitments (no "find their email" tasks)
 - [ ] State changes logged with `[Field → value]` notation
 - [ ] Bidirectional links are consistent
+- [ ] Requisition numbers detected and Role files created/enriched
+- [ ] Hiring manager inferred from calendar event organizers where applicable
+- [ ] Recruiter inferred from email CC fields where applicable
+- [ ] Domain lead resolved from hiring manager reporting chain where applicable
 - [ ] Graph state updated for processed files

package/template/.claude/skills/hiring-decision/SKILL.md

@@ -0,0 +1,366 @@
+---
+name: hiring-decision
+description: >
+  Synthesize all evidence (CV screening, interview assessments, transcripts)
+  into a final hiring recommendation. Produces a comprehensive recommendation
+  document with full evidence trail, level/track/discipline confirmation, and
+  a clear hire/no-hire decision. Use when all interview stages are complete.
+---
+
+# Hiring Decision
+
+Synthesize all available evidence for a candidate into a final hiring
+recommendation. This is the culmination of the hiring pipeline — every prior
+assessment feeds into this document.
+
+This is Stage 3 of a three-stage hiring pipeline:
+
+1. **Screen CV** — CV arrives → interview or pass
+2. **Assess Interview** — transcript arrives → updated evidence profile
+3. **Hiring Decision** (this skill) — all stages complete → hire or not
+
+## Trigger
+
+Run this skill:
+
+- When the user asks for a final hiring recommendation
+- When all planned interview stages for a candidate are complete
+- When the user asks "should we hire {Name}?"
+- When the user needs to compare finalists for a position
+
+## Prerequisites
+
+- `fit-pathway` CLI installed (`npx fit-pathway` must work)
+- Screening assessment: `knowledge/Candidates/{Name}/screening.md`
+- At least one interview assessment: `knowledge/Candidates/{Name}/interview-*.md`
+- Candidate brief: `knowledge/Candidates/{Name}/brief.md`
+- Transcripts provide additional context but are not strictly required if
+  interview assessments exist
+
+## Inputs
+
+- All candidate artifacts in `knowledge/Candidates/{Name}/`
+- Framework data via `fit-pathway`
+- `knowledge/Candidates/Insights.md` for cross-candidate context
+- `knowledge/Roles/*.md` — the Role file for this candidate's requisition
+  (provides remaining positions, hiring manager, domain lead priorities)
+- Other active candidates at the same level (for relative positioning)
+
+## Outputs
+
+- `knowledge/Candidates/{Name}/recommendation.md` — final hiring recommendation
+- Updated `knowledge/Candidates/{Name}/brief.md` — status and links
+- Updated `knowledge/Candidates/Insights.md` — cross-candidate observations
+
+---
+
+## Step 1: Gather All Evidence
+
+Collect every artifact for this candidate:
+
+```bash
+ls knowledge/Candidates/{Name}/
+```
+
+Read in order:
+
+1. `brief.md` — pipeline history, status, context
+2. `screening.md` — CV screening (Stage 1 output)
+3. `interview-*.md` — all interview assessments (Stage 2 outputs)
+4. `transcript-*.md` — raw transcripts for additional detail
+5. `panel.md` — panel brief if it exists (for continuity)
+
+Build a chronological evidence timeline:
+
+| Date | Stage | Source | Key Finding |
+|------|-------|--------|-------------|
+| {date} | CV Screening | screening.md | {key finding} |
+| {date} | {Interview type} | interview-{date}.md | {key finding} |
+
+## Step 2: Build Final Skill Profile
+
+For each skill in the target job's matrix, determine the **final rating** by
+selecting the highest-fidelity evidence available:
+
+**Evidence hierarchy** (highest to lowest fidelity):
+
+1. **Live demonstration** — candidate showed the skill in an interview exercise
+2. **Detailed interview discussion** — candidate gave specific, probing answers
+3. **Interviewer observation** — interviewer noted strength or concern
+4. **CV evidence with quantification** — concrete metrics and named projects
+5. **CV evidence without quantification** — described but vague
+6. **Not evidenced** — never surfaced across any stage
+
+```bash
+# Load the framework reference for final comparison
+npx fit-pathway job {discipline} {level} --track={track} --skills
+
+# Check progression if level is borderline
+npx fit-pathway progress {discipline} {level} --track={track}
+```
+
+For each skill, record:
+
+- **Final rating** — the proficiency level supported by the best evidence
+- **Best evidence source** — which stage provided the strongest signal
+- **Trajectory** — did the rating improve, decline, or hold across stages?
+- **Confidence** — high (demonstrated live), medium (discussed well), low
+  (CV only or thin evidence)
+
+## Step 3: Build Final Behaviour Profile
+
+For each framework behaviour, determine the **final maturity** using the same
+evidence hierarchy. Behaviours assessed in interviews carry far more weight
+than CV signals.
+
+```bash
+npx fit-pathway behaviour --list
+```
+
+For each behaviour:
+
+- **Final maturity** — the maturity level supported by interview evidence
+- **Best evidence** — specific moment or pattern across interviews
+- **Consistency** — was the behaviour shown once or throughout?
+
+## Step 4: Confirm Level and Track
+
+Using the complete evidence profile, make a final level and track recommendation:
+
+```bash
+# Compare adjacent levels
+npx fit-pathway job {discipline} {lower_level} --track={track}
+npx fit-pathway job {discipline} {target_level} --track={track}
+npx fit-pathway progress {discipline} {lower_level} --track={track}
+```
+
+| Question                                                    | Answer informs          |
+| ----------------------------------------------------------- | ----------------------- |
+| Does the candidate meet ≥ 70% of skills at the target level? | Level confirmation      |
+| Were level concerns from screening resolved in interviews?  | Level upgrade/downgrade |
+| Did interviewers explicitly suggest a different level?      | Strong level signal     |
+| Does the candidate's scope and autonomy match the level?    | Level fit               |
+| Which track energized the candidate in interviews?          | Track confirmation      |
+
+## Step 4b: Read Role Context
+
+If the candidate has a `Req` field in their brief, read the corresponding Role
+file:
+
+```bash
+ls knowledge/Roles/ | grep "{req_number}"
+cat "knowledge/Roles/{matching file}"
+```
+
+Extract and include in the recommendation:
+
+- **Remaining positions** on this req (from the Role file's `Positions` count
+  minus filled candidates in the Candidates table)
+- **Hiring manager** and their expectations (from the Role file and their People
+  note)
+- **Domain lead** and their hiring priorities (from recent meetings/emails)
+- **Other candidates** on the same req — how does this candidate compare to the
+  pipeline for this specific role?
+- **Channel** — is this a vendor candidate or HR candidate? This affects
+  onboarding timeline and engagement model.
+
+## Step 5: Assess Against Active Pipeline
+
+Check how this candidate compares to others at the same level:
+
+```bash
+# Read cross-candidate insights
+cat knowledge/Candidates/Insights.md
+
+# Check other candidates at the same level
+for dir in knowledge/Candidates/*/; do
+  if [ -f "$dir/screening.md" ]; then
+    head -5 "$dir/screening.md"
+  fi
+done
+```
+
+This is not a ranking exercise — it provides context. Note:
+
+- Whether this candidate fills a gap in the current pipeline
+- Whether stronger candidates exist for the same role
+- Whether this candidate is better suited to a different open position
+
+## Step 6: Write Hiring Recommendation
+
+Create `knowledge/Candidates/{Name}/recommendation.md`:
+
+```markdown
+# Hiring Recommendation — {Full Name}
+
+**Role:** {Discipline} {Level} — {Track}
+**Req:** {Req number and title, or "—"}
+**Hiring manager:** {Name from Role file, or "—"}
+**Domain lead:** {Name from Role file, or "—"}
+**Channel:** {hr / vendor}
+**Date:** {YYYY-MM-DD}
+**Prepared by:** Recruiter agent (with framework analysis)
+
+**⚠️ Advisory only — human decision required.**
+
+---
+
+## Recommendation
+
+**{Hire / Hire at {adjusted level} / Do not hire}**
+
+{3-5 sentence executive summary. Lead with the decision and the single
+strongest reason. Then address the primary risk and whether interviews
+resolved it. End with a clear statement of confidence.}
+
+---
+
+## Evidence Summary
+
+### Process Timeline
+
+| Date | Stage | Interviewer(s) | Outcome |
+|------|-------|----------------|---------|
+| {date} | CV Screening | — | {outcome} |
+| {date} | {Interview type} | {names} | {outcome} |
+
+### Final Skill Profile
+
+Framework reference: `{discipline} {level} --track={track}`
+
+| Skill | Expected | Final Rating | Confidence | Best Evidence | Trajectory |
+| --- | --- | --- | --- | --- | --- |
+| {skill} | {level} | {level} | {High/Med/Low} | {Stage that provided best evidence} | {↑ / ↓ / ―} |
+
+**Skill match:** {N}% Strong match, {N}% Adequate, {N}% Gap
+
+### Final Behaviour Profile
+
+| Behaviour | Expected | Final Maturity | Best Evidence |
+| --- | --- | --- | --- |
+| {behaviour} | {maturity} | {maturity} | {specific observation} |
+
+### Level Confirmation
+
+**Target level:** {original target}
+**Recommended level:** {confirmed or adjusted}
+
+{Paragraph explaining the level decision. Reference the framework progression
+criteria and specific evidence from interviews. If the level changed during
+the process, explain the journey (e.g. "Initially assessed at J100, screening
+identified scope concerns, decomposition confirmed J090 as the right fit").}
+
+### Track Confirmation
+
+**Recommended track:** {forward_deployed / platform / either}
+
+{Paragraph explaining track fit. Reference specific interview moments that
+revealed the candidate's natural orientation.}
+
+---
+
+## Decision Framework
+
+Apply these **decision rules** strictly:
+
+| Recommendation                | Criteria                                                                              |
+| ----------------------------- | ------------------------------------------------------------------------------------- |
+| **Hire**                      | ≥ 70% Strong match at final level, no unresolved core skill gaps, strong behaviours   |
+| **Hire at adjusted level**    | Strong candidate but evidence supports a different level than originally targeted      |
+| **Do not hire**               | Unresolved core skill gaps, behaviour concerns, or insufficient evidence after interviews |
+
+### Hire Criteria Check
+
+- [ ] ≥ 70% of skills rated Strong match at the recommended level
+- [ ] No core skill gaps remain unresolved (core = top-tier skills for the
+      discipline/track combination)
+- [ ] All framework behaviours at or above expected maturity
+- [ ] Level confirmed by interview evidence (not just CV)
+- [ ] Track fit confirmed by interview evidence
+- [ ] No red flags from any interview stage
+
+### Risk Assessment
+
+**Primary risk:** {The single biggest concern and whether it was resolved}
+**Mitigation:** {How the risk can be managed if hiring proceeds}
+
+**Secondary risks:**
+- {Risk 2 — severity and mitigation}
+
+---
+
+## Strengths to Leverage
+
+{What this person will bring to the team from day one. Frame in terms of
+business impact, not just technical capability.}
+
+1. **{Strength}:** {evidence and expected impact}
+2. **{Strength}:** {evidence and expected impact}
+
+## Development Areas
+
+{What this person will need to grow into. Frame as investment, not weakness.}
+
+1. **{Area}:** {current level → target level, suggested development path}
+2. **{Area}:** {current level → target level, suggested development path}
+
+---
+
+## Role Context
+
+{Context from the Role file to inform the hiring decision.}
+
+- **Requisition:** {Req number — title}
+- **Remaining positions:** {N of M filled}
+- **Hiring manager:** {name} | **Domain lead:** {name}
+- **Channel:** {hr / vendor — and implications for engagement model}
+
+## Pipeline Context
+
+{How this candidate compares to the active pipeline. Not a ranking — context
+for the hiring decision.}
+
+- **Pipeline position:** {Where this candidate sits relative to others on the
+  same requisition}
+- **Unique value:** {What this candidate offers that others in the pipeline don't}
+- **Alternative fit:** {If not hired for this role, could they fit another
+  open position? Reference other Role files.}
+
+---
+
+*This recommendation synthesizes all available evidence from {N} assessment
+stages conducted between {first date} and {last date}. The final decision
+rests with the hiring manager. All assessments are advisory.*
+```
+
+## Step 7: Update Candidate Brief and Insights
+
+Update `knowledge/Candidates/{Name}/brief.md`:
+
+- Update **Status** to reflect the recommendation (`recommended` / `not-recommended`)
+- Add link: `- [Hiring Recommendation](./recommendation.md)`
+- Add final pipeline entry with the recommendation date and outcome
+
+Update `knowledge/Candidates/Insights.md` with any cross-candidate observations:
+
+- If this candidate is the strongest at their level, note it
+- If this candidate revealed a pattern about a sourcing channel, note the channel
+- If the level adjustment has implications for other candidates, note it
+
+**Use precise edits — don't rewrite entire files.**
+
+## Quality Checklist
+
+- [ ] Every skill rating traces to a specific evidence source (stage + detail)
+- [ ] Evidence hierarchy is respected — interview evidence outranks CV evidence
+- [ ] Level recommendation is grounded in framework progression criteria
+- [ ] Track recommendation cites interview evidence, not just CV signals
+- [ ] Decision rules are applied strictly — verify percentages and gap counts
+- [ ] Risk assessment is honest — don't minimize real concerns to justify hiring
+- [ ] Development areas are specific and actionable
+- [ ] Pipeline context is factual — no ranking by protected characteristics
+- [ ] Cross-candidate insights added to Insights.md where relevant
+- [ ] Brief updated with recommendation link and status
+- [ ] "Do not hire" is explained with the same rigour as "Hire" — the candidate
+      deserves to know why if they request access (GDPR Article 15)
+- [ ] Recommendation could be shown to the candidate without embarrassment

package/template/.claude/skills/meeting-prep/SKILL.md

@@ -25,6 +25,8 @@ meetings.
 - `knowledge/People/*.md` — attendee context
 - `knowledge/Organizations/*.md` — company context
 - `knowledge/Projects/*.md` — project context
+- `knowledge/Candidates/*/brief.md` — candidate context (for interview meetings)
+- `knowledge/Roles/*.md` — role/requisition context (for interview meetings)
 
 ## Outputs
 
@@ -130,6 +132,40 @@ Suggested Talking Points
 - Talking points: concrete, not generic
 - If no notes exist for a person, mention that and offer to create one
 
+### Interview Meeting Context
+
+When preparing for interview meetings (title contains "Interview", "Screening",
+"Decomposition", "Panel", or a candidate name from `knowledge/Candidates/`):
+
+1. **Read the candidate brief:** `knowledge/Candidates/{Name}/brief.md`
+2. **Read the Role file:** Look up the `Req` field and read the corresponding
+   `knowledge/Roles/*.md` file.
+3. **Include in the briefing:**
+   - Candidate's current status, skills, and screening recommendation
+   - Role context: hiring manager, domain lead, remaining positions
+   - Other candidates on the same requisition and their statuses (from the
+     Role file's Candidates table)
+   - Previous interview assessments if this is a second/later stage
+   - Panel brief if one exists (`panel.md`)
+4. **Format as a dedicated section:**
+
+```markdown
+## Candidate: {Name}
+**Role:** {Title from Role file} | **Req:** {req number}
+**Hiring manager:** {name} | **Domain lead:** {name}
+**Status:** {current pipeline status}
+**Screening:** {Interview / Interview with focus areas / Pass}
+
+### Pipeline for this req
+- {N} candidates total, {N} interviewed, {N} remaining positions
+
+### Key strengths
+- {from screening.md}
+
+### Focus areas for this interview
+- {from screening.md or panel.md}
+```
+
 ## Constraints
 
 - Only prep for meetings with external attendees

package/template/.claude/skills/right-to-be-forgotten/SKILL.md

@@ -241,7 +241,7 @@ Create the audit trail at `knowledge/Erasure/{Name}--{YYYY-MM-DD}.md`:
 ### Deleted Files
 - `knowledge/Candidates/{Name}/brief.md`
 - `knowledge/Candidates/{Name}/CV.pdf`
-- `knowledge/Candidates/{Name}/assessment.md`
+- `knowledge/Candidates/{Name}/screening.md`
 - `knowledge/People/{Name}.md`
 - {list all deleted files}