@esoteric-logic/praxis-harness 3.3.1 → 3.5.0

package/base/skills/px-prompt/SKILL.md

@@ -39,15 +39,14 @@ Do NOT show "Work on praxis" or "Work on maximus" or any project name — the en
 
 Platform outputs use suffixed filenames so users can distinguish them at a glance:
 
-| Platform | Output Filename | Budget | When |
-|----------|----------------|--------|------|
-| Claude Projects | `system-prompt.md` | 5,000 chars | Standalone projects |
-| Claude Projects | `project-instructions-claude-desktop.md` | 5,000 chars | Compiled projects (compiler output) |
-| Perplexity Spaces | `space-instructions-perplexity.md` | 4,000 chars | All projects |
+| Platform | Output Filename | Budget |
+|----------|----------------|--------|
+| Claude Projects | `system-prompt.md` | 5,000 chars |
+| Perplexity Spaces | `space-instructions-perplexity.md` | 4,500 chars |
 
-Each project has exactly **2 output files** — one for Claude Projects, one for Perplexity Spaces.
-- Standalone: `system-prompt.md` + `space-instructions-perplexity.md`
-- Compiled: `project-instructions-claude-desktop.md` + `space-instructions-perplexity.md`
+Each project has exactly **2 output files** — one per platform, same filenames regardless of mode.
+- `system-prompt.md` — standalone: hand-written, compiled: generated from profile blocks
+- `space-instructions-perplexity.md` — standalone: condensed from system-prompt, compiled: assembled from blocks
 
 Claude Code `CLAUDE.md` is NOT generated by this skill — use `px-scaffold`.
 
@@ -371,7 +370,7 @@ Read the full `system-prompt.md` as source.
 
 ### 3a. Generate Perplexity Space instructions
 
-**Target:** `space-instructions-perplexity.md` | **Budget:** under 4,000 chars
+**Target:** `space-instructions-perplexity.md` | **Budget:** under 4,500 chars
 
 **Include:** identity, domain expertise, research domains, source priority, answer format, key frameworks (by name only), reasoning approach, accuracy standards, anti-hallucination rules
 **Exclude:** internal templates, scoring matrices, reference file content, deployment details, full tables
@@ -416,7 +415,7 @@ Think step-by-step: Understand the question → search sources → analyze findi
 
 After generating, check:
 - `system-prompt.md` under 5,000 chars
-- `space-instructions-perplexity.md` under 4,000 chars
+- `space-instructions-perplexity.md` under 4,500 chars
 
 If over budget: flag and suggest sections to trim.
 
@@ -447,7 +446,7 @@ node bin/prompt-compile.js <project-name>
 
 **Claude Desktop / Projects (claude.ai):**
 1. Open project at claude.ai/projects → "Set project instructions"
-2. Paste `system-prompt.md` (standalone) or `project-instructions-claude-desktop.md` (compiled)
+2. Paste `system-prompt.md`
 3. If `references/` exists: upload each `.md` file as project knowledge
 4. Save
 
@@ -493,7 +492,7 @@ Print deployment reminders for any project with changes.
 ### 6a. Read all project files
 1. Read `prompt-config.yaml` for project metadata
 2. Read `system-prompt.md` (source of truth)
-3. Read platform outputs (`system-prompt.md` or `project-instructions-claude-desktop.md`, `space-instructions-perplexity.md`)
+3. Read platform outputs (`system-prompt.md`, `space-instructions-perplexity.md`)
 4. Read all files in `references/` directory
 5. List any other files in the project folder
 
@@ -509,9 +508,8 @@ Check each file against these criteria:
 - Are all referenced knowledge files present in references/?
 
 **Budget checks:**
-- `system-prompt.md` under 5,000 chars? (standalone)
-- `project-instructions-claude-desktop.md` under 5,000 chars? (compiled)
-- `space-instructions-perplexity.md` under 4,000 chars?
+- `system-prompt.md` under 5,000 chars?
+- `space-instructions-perplexity.md` under 4,500 chars?
 
 **Currency checks (via Perplexity):**
 - Are domain-specific terms, standards, and versions still current?
@@ -524,7 +522,7 @@ Show a structured report:
 |---------------------------|--------|-----------------------------------|
 | Quality Controls section  | PASS   | Present in system-prompt.md       |
 | Anti-Hallucination        | FAIL   | Missing — will add                |
-| Budget: Perplexity        | PASS   | 2,392 / 4,000 chars              |
+| Budget: Perplexity        | PASS   | 2,392 / 4,500 chars              |
 | File naming               | FAIL   | Uses old convention — will rename |
 | Knowledge files           | WARN   | 0 reference files — consider adding |
 ```
@@ -737,7 +735,7 @@ Maximus PP:  2 contracts at IRS (from USASpending)
 Gaps:        Key personnel [RESEARCH NEEDED]
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 Files created:
-  ✓ project-instructions-claude-desktop.md (2,480 chars) — compiled from profile
+  ✓ system-prompt.md (2,480 chars) — compiled from profile
   ✓ space-instructions-perplexity.md (3,976 chars) — deal-specific research domains
   ✓ references/irs-masterfile-intel.md
   ✓ knowledge/deal-context.md
@@ -930,7 +928,7 @@ Determine which platforms are targets from `prompt-config.yaml`.
 ### 11b. Deploy sequence (per platform)
 
 **For Claude Projects / Desktop:**
-1. Determine file: `system-prompt.md` (standalone) or `project-instructions-claude-desktop.md` (compiled)
+1. Read `system-prompt.md`
 2. Copy to clipboard: `cat <file> | pbcopy`
 3. Print: "Copied to clipboard. Paste at: claude.ai/projects → Set project instructions"
 4. List knowledge files to upload: all `.md` files from `references/` AND `knowledge/`
@@ -1102,8 +1100,7 @@ NEXT GATE CHECKLIST: Pre-Proposal
 - **Claude Code CLAUDE.md**: NOT generated by px-prompt — use px-scaffold
 
 ### File naming
-- Standalone Claude Projects output: `system-prompt.md`
-- Compiled Claude Projects output: `project-instructions-claude-desktop.md`
+- Claude Projects output: `system-prompt.md` (all modes)
 - Perplexity output: `space-instructions-perplexity.md`
 
 ### Quality defaults (mandatory in all generated prompts)

package/bin/prompt-compile.js

@@ -58,7 +58,7 @@ function discoverAllProjects() {
 const CHAR_BUDGETS = {
   'claude-code': Infinity,
   'claude-project': 5000,
-  'perplexity-space': 4000,
+  'perplexity-space': 4500,
 };
 
 // Global flags set by CLI parser
@@ -213,7 +213,7 @@ function compileProject(projectName, targets, projectDirOverride, clientDirOverr
 
   const outputNames = {
     'claude-code': 'CLAUDE.md',
-    'claude-project': 'project-instructions-claude-desktop.md',
+    'claude-project': 'system-prompt.md',
     'perplexity-space': 'space-instructions-perplexity.md',
   };
 
@@ -302,8 +302,8 @@ function main() {
       console.log('No projects found.');
       process.exit(0);
     }
-    console.log(`${'Project'.padEnd(28)} ${'Mode'.padEnd(12)} ${'System Prompt'.padEnd(15)} ${'Claude Desktop'.padEnd(15)} ${'Perplexity'.padEnd(15)} Refs`);
-    console.log('-'.repeat(105));
+    console.log(`${'Project'.padEnd(28)} ${'Mode'.padEnd(12)} ${'System Prompt'.padEnd(15)} ${'Perplexity'.padEnd(15)} Refs`);
+    console.log('-'.repeat(85));
     for (const proj of allProjects) {
       const cfgPath = path.join(proj.dir, 'prompt-config.yaml');
       const cfg = fs.existsSync(cfgPath) ? yaml.load(fs.readFileSync(cfgPath, 'utf8')) : {};
@@ -318,7 +318,7 @@ function main() {
         ? fs.readdirSync(refsDir).filter((f) => f.endsWith('.md')).length
         : 0;
       console.log(
-        `${proj.name.padEnd(28)} ${mode.padEnd(12)} ${fileStatus('system-prompt.md').padEnd(15)} ${fileStatus('project-instructions-claude-desktop.md').padEnd(15)} ${fileStatus('space-instructions-perplexity.md').padEnd(15)} ${refCount}`
+        `${proj.name.padEnd(28)} ${mode.padEnd(12)} ${fileStatus('system-prompt.md').padEnd(15)} ${fileStatus('space-instructions-perplexity.md').padEnd(15)} ${refCount}`
       );
     }
     process.exit(0);
@@ -338,7 +338,7 @@ function main() {
     console.log('\n\x1b[1mPROMPT ENGINE DASHBOARD\x1b[0m');
     console.log('\x1b[90m' + '━'.repeat(110) + '\x1b[0m');
     console.log(
-      `${'Project'.padEnd(24)} ${'Mode'.padEnd(12)} ${'Claude Proj'.padEnd(14)} ${'Perplexity'.padEnd(14)} ${'Refs'.padEnd(6)} ${'Updated'.padEnd(12)} Stale?`
+      `${'Project'.padEnd(24)} ${'Mode'.padEnd(12)} ${'System Prompt'.padEnd(14)} ${'Perplexity'.padEnd(14)} ${'Refs'.padEnd(6)} ${'Updated'.padEnd(12)} Stale?`
     );
     console.log('\x1b[90m' + '─'.repeat(110) + '\x1b[0m');
 
@@ -371,7 +371,7 @@ function main() {
       const stale = daysSince > STALE_DAYS ? '\x1b[31mYes\x1b[0m' : '\x1b[32mNo\x1b[0m';
 
       console.log(
-        `${proj.name.padEnd(24)} ${mode.padEnd(12)} ${fileBudget('project-instructions-claude-desktop.md', CHAR_BUDGETS['claude-project']).padEnd(23)} ${fileBudget('space-instructions-perplexity.md', CHAR_BUDGETS['perplexity-space']).padEnd(23)} ${String(refCount).padEnd(6)} ${updated.padEnd(12)} ${stale}`
+        `${proj.name.padEnd(24)} ${mode.padEnd(12)} ${fileBudget('system-prompt.md', CHAR_BUDGETS['claude-project']).padEnd(23)} ${fileBudget('space-instructions-perplexity.md', CHAR_BUDGETS['perplexity-space']).padEnd(23)} ${String(refCount).padEnd(6)} ${updated.padEnd(12)} ${stale}`
       );
     }
 

package/lib/assemblers.js

@@ -148,6 +148,20 @@ function assembleClaudeProject(blocks, projectConfig, vars) {
     for (const block of formatBlocks) lines.push(block.content, '');
   }
 
+  // Deal context — auto-injected from deal config vars
+  if (projectConfig.deal && vars.agency) {
+    lines.push('## Deal Context');
+    lines.push(`**Customer**: ${vars.agency}`);
+    if (vars.program_name) lines.push(`**Program**: ${vars.program_name}`);
+    if (vars.incumbent) lines.push(`**Incumbent**: ${vars.incumbent}`);
+    if (vars.contract_value) lines.push(`**Value**: ${vars.contract_value}`);
+    if (vars.naics) lines.push(`**NAICS**: ${vars.naics}`);
+    if (vars.pop) lines.push(`**POP**: ${vars.pop}`);
+    if (projectConfig.deal_type) lines.push(`**Type**: ${projectConfig.deal_type}`);
+    if (projectConfig.capture_phase) lines.push(`**Phase**: ${projectConfig.capture_phase}`);
+    lines.push('');
+  }
+
   // Additional context
   const append = (projectConfig.overrides || {}).claude_project_append || {};
   if (append.additional_context) {
@@ -222,6 +236,18 @@ function assemblePerplexitySpace(blocks, projectConfig, vars) {
     lines.push('');
   }
 
+  // Deal context — auto-injected from deal config vars
+  if (projectConfig.deal && vars.agency) {
+    lines.push('## Deal Context');
+    lines.push(`Customer: ${vars.agency}`);
+    if (vars.program_name) lines.push(`Program: ${vars.program_name}`);
+    if (vars.incumbent) lines.push(`Incumbent: ${vars.incumbent}`);
+    if (vars.contract_value) lines.push(`Value: ${vars.contract_value}`);
+    if (projectConfig.deal_type) lines.push(`Type: ${projectConfig.deal_type}`);
+    if (projectConfig.capture_phase) lines.push(`Phase: ${projectConfig.capture_phase}`);
+    lines.push('');
+  }
+
   // Research Domains
   const append = (projectConfig.overrides || {}).perplexity_space_append || {};
   if (append.research_domains) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@esoteric-logic/praxis-harness",
-  "version": "3.3.1",
+  "version": "3.5.0",
   "description": "Layered Claude Code harness — workflow discipline, AI-Kits, persistent vault integration",
   "bin": {
     "praxis-harness": "./bin/praxis.js",

package/prompts/work/maximus/projects/benefeds/space-instructions-perplexity.md

@@ -12,6 +12,14 @@ SSEB simulation: score only what's on the page against Section M. Ratings: Outst
 11-section assessment (Customer/Architecture/Process/Artifacts/Planning/Assumptions/Risks/Dependencies/Cyber/Cost/Competitive). PAMASI maturity: P→A→M→A→S→I. Gates: Shaping=P, Mid Capture=A-M, Pre-Proposal=S, Pre-Submission=I.
 BLUF every paragraph. FBP every claim. Active voice, SHALL→WILL, 70/30 mission/company. Ban "robust/world-class/seamless/leverage/synergy." Hierarchy: Approach→Framework→Methodology→Process (never conflate).
 
+## Deal Context
+Customer: Office of Personnel Management (OPM)
+Program: BENEFEDS
+Incumbent: FedPoint Systems LLC (Long Term Care Partners / John Hancock)
+Value: $500M ceiling
+Type: recompete
+Phase: Shaping
+
 ## Research Domains
 - OPM BENEFEDS contract history and performance (SAM.gov, FPDS)
 - FedPoint federal performance and ATO certifications

package/prompts/work/maximus/projects/benefeds/{project-instructions-claude-desktop.md → system-prompt.md}

@@ -46,6 +46,16 @@ Match response length to question complexity. Lead with the answer. No preamble
 
 11-section scorecard (R/Y/G) + PAMASI stage + phase verdict (On Track/Needs Work/Off Track). Gate verdicts: Pass/Conditional/No Pass/Stop & Reset.
 
+## Deal Context
+**Customer**: Office of Personnel Management (OPM)
+**Program**: BENEFEDS
+**Incumbent**: FedPoint Systems LLC (Long Term Care Partners / John Hancock)
+**Value**: $500M ceiling
+**NAICS**: 518210, 524292, 541512
+**POP**: Jan 2023 – Dec 2026
+**Type**: recompete
+**Phase**: Shaping
+
 ## Knowledge Files
 Upload these alongside this prompt:
 - **references/maturity-questions.md** — maturity-questions

package/prompts/work/maximus/projects/dha-tricare/prompt-config.yaml

@@ -10,7 +10,13 @@ platforms:
   - claude-project
   - perplexity-space
 
-vars: {}
+vars:
+  agency: "Defense Health Agency (DHA)"
+  program_name: "TRICARE Managed Care Support / Health IT Services"
+  incumbent: "Humana Government Business (East), TriWest Healthcare Alliance (West), Leidos (MHS GENESIS)"
+  contract_value: "T-5: $70.9B (East) + $65.1B (West) + $6.9B (MHS GENESIS)"
+  naics: "524114"
+  pop: "T-5 active through ~2032. T-6 recompete expected ~2031."
 
 knowledge_files:
   - file: references/dha-tricare-intel.md
@@ -33,8 +39,8 @@ knowledge_packs:
 overrides:
   perplexity_space_append:
     research_domains: |
-      - DHA TRICARE T-5/T-6 contract developments (SAM.gov, USASpending, FPDS)
-      - DHA strategic plan, IG/GAO findings, congressional oversight
-      - TRICARE beneficiary experience, claims processing, provider networks
-      - MHS GENESIS modernization and cloud migration
-      - Maximus federal health IT and CX past performance
+      - DHA TRICARE T-5/T-6 contract developments (SAM.gov, FPDS)
+      - DHA strategic plan, GAO findings, congressional oversight
+      - TRICARE beneficiary experience and provider networks
+      - MHS GENESIS modernization
+      - Maximus federal health IT past performance

package/prompts/work/maximus/projects/dha-tricare/space-instructions-perplexity.md

@@ -12,12 +12,20 @@ SSEB simulation: score only what's on the page against Section M. Ratings: Outst
 11-section assessment (Customer/Architecture/Process/Artifacts/Planning/Assumptions/Risks/Dependencies/Cyber/Cost/Competitive). PAMASI maturity: P→A→M→A→S→I. Gates: Shaping=P, Mid Capture=A-M, Pre-Proposal=S, Pre-Submission=I.
 BLUF every paragraph. FBP every claim. Active voice, SHALL→WILL, 70/30 mission/company. Ban "robust/world-class/seamless/leverage/synergy." Hierarchy: Approach→Framework→Methodology→Process (never conflate).
 
+## Deal Context
+Customer: Defense Health Agency (DHA)
+Program: TRICARE Managed Care Support / Health IT Services
+Incumbent: Humana Government Business (East), TriWest Healthcare Alliance (West), Leidos (MHS GENESIS)
+Value: T-5: $70.9B (East) + $65.1B (West) + $6.9B (MHS GENESIS)
+Type: new-start
+Phase: Shaping
+
 ## Research Domains
-- DHA TRICARE T-5/T-6 contract developments (SAM.gov, USASpending, FPDS)
-- DHA strategic plan, IG/GAO findings, congressional oversight
-- TRICARE beneficiary experience, claims processing, provider networks
-- MHS GENESIS modernization and cloud migration
-- Maximus federal health IT and CX past performance
+- DHA TRICARE T-5/T-6 contract developments (SAM.gov, FPDS)
+- DHA strategic plan, GAO findings, congressional oversight
+- TRICARE beneficiary experience and provider networks
+- MHS GENESIS modernization
+- Maximus federal health IT past performance
 
 ## How to Answer
 No flattery or filler. Be skeptical, concise.

package/prompts/work/maximus/projects/dha-tricare/{project-instructions-claude-desktop.md → system-prompt.md}

@@ -46,6 +46,16 @@ Match response length to question complexity. Lead with the answer. No preamble
 
 11-section scorecard (R/Y/G) + PAMASI stage + phase verdict (On Track/Needs Work/Off Track). Gate verdicts: Pass/Conditional/No Pass/Stop & Reset.
 
+## Deal Context
+**Customer**: Defense Health Agency (DHA)
+**Program**: TRICARE Managed Care Support / Health IT Services
+**Incumbent**: Humana Government Business (East), TriWest Healthcare Alliance (West), Leidos (MHS GENESIS)
+**Value**: T-5: $70.9B (East) + $65.1B (West) + $6.9B (MHS GENESIS)
+**NAICS**: 524114
+**POP**: T-5 active through ~2032. T-6 recompete expected ~2031.
+**Type**: new-start
+**Phase**: Shaping
+
 ## Knowledge Files
 Upload these alongside this prompt:
 - **references/maturity-questions.md** — maturity-questions