Compare commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
c103def91b | ||
|
|
485df6bd0d | ||
|
|
86bdca5f22 |
No files matched your search
@@ -22,6 +22,16 @@ change's documentation disposition and the evidence behind its merge; commands
|
||||
presented as runnable must be ready to run as written. See
|
||||
`docs/agents/standards.md`.
|
||||
|
||||
### Credentials
|
||||
|
||||
When T3 Code's `request_secret` tool is available, call it whenever an
|
||||
authorized task needs a sudo password or other credential from the user.
|
||||
Use it instead of asking the user to paste secrets, copy sudo commands,
|
||||
or enter credentials in a terminal. For shell commands use `delivery: "file"`
|
||||
and read the private file inside the command; for supported T3 tools pass
|
||||
`secretRef`. Keep secret values out of output and execute the authorized
|
||||
operation yourself. Pass this rule to delegated agents.
|
||||
|
||||
### Focused checks
|
||||
|
||||
`python3 -m unittest discover -s tests -v`
|
||||
|
||||
@@ -101,3 +101,5 @@ artifacts, verification and Operator acceptance. These packages customize
|
||||
answer/detail/example/acknowledgement presentation only.
|
||||
|
||||
The automatic `research-auto` Planner package supplies primary-source selection, preferred-domain exploration/refinement and evidence-reading methodology, publication/version checking, corroboration and already-admitted empirical evidence methods. The manual `research` package remains available on request. Tool arguments, budgets, continuation, recovery and untrusted research provenance belong to the owner.
|
||||
|
||||
Planner direct-address presentation is supplied by planner-communication v0.14.0. Other packages retain their existing versions. Forks can replace this presentation while retaining communication_context treatment metadata.
|
||||
@@ -3,9 +3,11 @@ name: planner-communication
|
||||
description: Communication style for the assigned planner; proportional answers, examples, Markdown and natural acknowledgements.
|
||||
metadata:
|
||||
relayview-display-name: Planner communication
|
||||
relayview-version: 0.13.0
|
||||
relayview-version: 0.14.0
|
||||
---
|
||||
|
||||
Address the person you are speaking with as “you” and describe their decisions, permissions, and next actions from their perspective. Explain workflow behavior in familiar language, naming the responsible agent or RelayView when the software acts. Before publishing a reply, check that the person can tell who acts next and what, if anything, you need from them.
|
||||
|
||||
Lead with the answer, decision, observed finding, or material uncertainty. Use detail proportional to the immediate question, including requested examples and technical mechanics, with no fixed word, line, or token limit. Keep inspection effort independent of conversational length.
|
||||
|
||||
For user-facing focused Chat messages, use Markdown when it makes the answer easier to read. A simple reply may be plain prose. Use concise headings, lists, tables, and inline code only when they help the content; avoid decorative structure. When code needs a block, use a fenced code block with the correct language identifier if the language is known. Leave a fence unlabeled when the language is unknown or the content is genuinely plain.
|
||||
|
||||
@@ -2,7 +2,7 @@
|
||||
"schemaVersion": 3,
|
||||
"source": {
|
||||
"kind": "first-party",
|
||||
"reference": "v0.13.0"
|
||||
"reference": "v0.14.0"
|
||||
},
|
||||
"role": "planner",
|
||||
"step": "planning",
|
||||
@@ -10,7 +10,7 @@
|
||||
"files": [
|
||||
{
|
||||
"path": "SKILL.md",
|
||||
"sha256": "fe45e08225328bab250475bc966bf480320459320d70a2be6dc07655eda37cc6"
|
||||
"sha256": "26154c0d1985dc0214fde7c42f02e0581c4779c2e41b027d676acf5b0b23edfc"
|
||||
}
|
||||
]
|
||||
}
|
||||
Reference in new issue
Block a user