parse-reply

name	Parse Reply
description	Extract corrections from user's reply to a brief email.

Input

Text of the user's reply to a brief email.

Pre-processing

1. Strip HTML (if input is HTML)

If the input contains HTML tags, extract plain text before doing anything else:

• Remove all HTML tags (e.g. <div>, <p>, <br>, <span>, <b>, <blockquote>, etc.)
• Decode HTML entities (  → space, & → &, < → <, > → >, " → ")
• Collapse multiple consecutive blank lines into a single blank line
• Strip leading/trailing whitespace from each line

2. Strip quoted original email

After HTML stripping (or if input was already plain text), remove everything below the first occurrence of any of these separator patterns:

• A line starting with From:
• A line starting with Sent:
• A line that is exactly --- or ___ or ***
• A line starting with > followed by On ... wrote:
• A line starting with ________________________________ (Outlook's horizontal rule in plain-text format)
• A line matching On .{5,80} wrote: (Outlook's "On Mon 24 Mar, Dayarc <...> wrote:" pattern)

Only parse the text above the first separator.

3. Strip email signatures

After removing quoted text, also discard lines from the first occurrence of a signature block marker:

• A line that is exactly -- (standard signature delimiter)
• A line containing only Sent from my iPhone / Sent from my Android / Get Outlook for iOS (and similar auto-appended footers)

Only parse the text above the first signature marker.

Output (JSON)

{
  "corrections": [{
    "action": "mark_done",
    "target": "auth migration",
    "detail": "User says auth migration is complete"
  }],
  "quality_signals": [{
    "sentiment": "negative",
    "detail": "too much noise in inbox section"
  }]
}

Instructions

The user may write in any free-form style — there are no required keywords or formats. Read each sentence for intent and extract meaning, even when phrasing is indirect or implicit.

1. Parse natural language corrections. Infer the action from context — do not require specific trigger words:
   • mark_done: any sentence conveying that something was completed, sent, fixed, resolved, or handled. Examples of intent (not exhaustive): "X is done", "finished X", "I already fixed the issue related to X", "I already sent the mail to my peer about X", "that's shipped", "merged", "resolved the X problem".
   • remove: any sentence conveying that something is no longer relevant, was cancelled, doesn't need tracking, or should be dropped. Examples: "drop X", "X got cancelled", "no longer relevant", "ignore X", "we decided not to do X".
   • add_priority: any sentence conveying a new thing to track, do, or remember. Examples: "remind me to X", "I need to X", "add X as priority", "don't forget X", "new priority: X", "I should chat with PM".
   • correct: any sentence conveying that existing information is wrong or needs updating. Examples: "actually X is ...", "correction: ...", "X should be Y not Z".
   • For indirect references (e.g. "I already fixed the issue related to the UI"): extract the most specific identifiable subject as target (e.g. "UI issue") and note the indirect phrasing in detail.
2. Parse quality signals — sentences that evaluate the brief itself (not specific work items). Classify sentiment and extract detail:
   • positive: any expression of approval, usefulness, or accuracy about the brief. Examples: "great brief", "spot on", "exactly right", "really useful", "perfect", "loved the summary".
   • negative: any expression that the brief was noisy, inaccurate, incomplete, or off-focus. Examples: "too much noise", "priorities were off", "missed my X work", "not relevant", "wrong focus", "too long", "cluttered".
   • Map each match to { sentiment: "positive"|"negative", detail: "<normalized phrase or quoted fragment>" }.
3. Ignore purely social/phatic replies with no work content: "thanks", "looks good", "ok", "got it", "👍". These produce neither a correction nor a quality signal.
4. A reply may contain both corrections and quality signals — extract all.
5. If no actionable content found, return empty arrays for both corrections and quality_signals.

Examples

Single correction:

Input: "Auth migration is done, shipped yesterday"
Output: { "corrections": [{ "action": "mark_done", "target": "auth migration", "detail": "Shipped yesterday" }], "quality_signals": [] }

Multiple corrections:

Input: "Drop the perf review prep — it got cancelled. Also add quarterly OKR planning as a new priority."
Output: {
  "corrections": [
    { "action": "remove", "target": "perf review prep", "detail": "Cancelled" },
    { "action": "add_priority", "target": "quarterly OKR planning", "detail": "New priority from user" }
  ],
  "quality_signals": []
}

Quoted reply (strip original):

Input:
"Cost optimization is done

From: Dayarc <[e-mail]>
Sent: Monday, March 24, 2026 8:00 PM
Subject: 🌙 PM Brief — Mon 24 Mar

Section A: What you accomplished today
..."

Output: { "corrections": [{ "action": "mark_done", "target": "cost optimization", "detail": "User confirmed done" }], "quality_signals": [] }

Quality signal (positive):

Input: "Great brief today, really useful"
Output: { "corrections": [], "quality_signals": [{ "sentiment": "positive", "detail": "great brief, really useful" }] }

Quality signal (negative):

Input: "Too much noise in the inbox section, priorities were off"
Output: { "corrections": [], "quality_signals": [{ "sentiment": "negative", "detail": "too much noise in inbox section" }, { "sentiment": "negative", "detail": "priorities were off" }] }

Mixed correction and quality signal:

Input: "Auth migration is done. Also the brief missed my PR review work."
Output: {
  "corrections": [{ "action": "mark_done", "target": "auth migration", "detail": "User confirmed done" }],
  "quality_signals": [{ "sentiment": "negative", "detail": "missed PR review work" }]
}

No-op reply:

Input: "Looks good, thanks!"
Output: { "corrections": [], "quality_signals": [] }

Non-actionable single-word/emoji reply:

Input: "ok"
Output: { "corrections": [], "quality_signals": [] }

Input: "👍"
Output: { "corrections": [], "quality_signals": [] }

Free-form implicit completion (indirect reference to a work item):

Input: "I already fixed the issue that related to the UI"
Output: { "corrections": [{ "action": "mark_done", "target": "UI issue", "detail": "User says already fixed" }], "quality_signals": [] }

Free-form implicit completion (action already taken):

Input: "I already sent the mail to my peer"
Output: { "corrections": [{ "action": "mark_done", "target": "send mail to peer", "detail": "User says already sent" }], "quality_signals": [] }

Free-form reminder / new priority:

Input: "remind me to chat with PM"
Output: { "corrections": [{ "action": "add_priority", "target": "chat with PM", "detail": "User reminder" }], "quality_signals": [] }

Mixed free-form reply:

Input: "I already fixed the UI issue. Also remind me to chat with PM about the roadmap. The brief was a bit too long though."
Output: {
  "corrections": [
    { "action": "mark_done", "target": "UI issue", "detail": "User says already fixed" },
    { "action": "add_priority", "target": "chat with PM about roadmap", "detail": "User reminder" }
  ],
  "quality_signals": [{ "sentiment": "negative", "detail": "brief was too long" }]
}

HTML-wrapped reply with signature (Outlook format):

Input:
"<html><body><div style='font-family:Calibri'>
<p>Auth migration is done.</p>
<p>Also drop the perf review prep item.</p>
<div>-- <br>Jane Smith<br>Engineering Manager</div>
<blockquote style='border-left:1px solid #ccc'>
<div>From: Dayarc &lt;[e-mail]&gt;</div>
<div>Sent: Tuesday, April 9, 2026 8:00 PM</div>
<div>Subject: 🌙 Evening Wrap-up — Tue 9 Apr</div>
<div>... original brief ...</div>
</blockquote>
</div></body></html>"

Output: {
  "corrections": [
    { "action": "mark_done", "target": "auth migration", "detail": "User confirmed done" },
    { "action": "remove", "target": "perf review prep", "detail": "User requested removal" }
  ],
  "quality_signals": []
}

HTML reply with Outlook horizontal rule separator:

Input:
"<html><body><div>
<p>Cost optimization review is complete.</p>
<p>________________________________</p>
<p>From: Dayarc &lt;[e-mail]&gt;<br>
Sent: Wednesday, April 9, 2026<br>
Subject: ☀️ Morning Brief — Wed 9 Apr</p>
</div></body></html>"

Output: { "corrections": [{ "action": "mark_done", "target": "cost optimization review", "detail": "User confirmed complete" }], "quality_signals": [] }

All five correction types in one reply:

Input: "Auth migration is done. Drop the perf review — it got cancelled. Add quarterly OKR planning as priority. Actually the ARM CI issue is a flaky test, not a build failure."
Output: {
  "corrections": [
    { "action": "mark_done", "target": "auth migration", "detail": "User confirmed done" },
    { "action": "remove", "target": "perf review", "detail": "Cancelled" },
    { "action": "add_priority", "target": "quarterly OKR planning", "detail": "New priority from user" },
    { "action": "correct", "target": "ARM CI issue", "detail": "Flaky test, not a build failure" }
  ],
  "quality_signals": []
}

---

Dry-Run Validation

When running in conversational (non-scheduled) mode, after returning the parsed JSON, echo a human-readable summary to the terminal:

📬 Reply parsed:
  Corrections (3):
    ✅ mark_done  → auth migration
    ❌ remove     → perf review prep
    ➕ add_priority → quarterly OKR planning
  Quality signals (1):
    👎 negative: "priorities were off"
  Profile fields that will be updated:
    active_threads: auth migration → status: "done"
    priorities_today: perf review prep removed
    priorities_today: quarterly OKR planning added (urgency: 🟡 soon)
    feedback: { sentiment: "negative", detail: "priorities were off" }

If corrections is empty and quality_signals is empty, output:

📬 Reply parsed: no actionable content — skipping profile update.

This dry-run echo happens before any memory write, so the user can verify corrections are correct before they're applied.